veals interesting cultural differences between field- based pale-
ontologists, who, inured to the idiosyncrasies of fossil life, are
willing to embrace the idea of nonsteady rates of evolution,
versus lab- based molecular biologists, who see mechanism
in cellular structure and are more orthodox uniformitarians
than their geologic counterparts. While the Precambrian is by
no means the obscure, unmapped expanse it was to Victorian
geologists, the transition from it across the threshold into the
Cambrian remains dimly lit.
In most paleontologic textbooks, the Cambrian explosion
is the start of the story, the prelude to the rollicking tale of
trilobites, lungfish, coal swamps, tyrannosaurs, pterodactyls,
megatheria, mammoths, and hominids. In the most important
ways, though, the Cambrian world was not so different from
116 Ch a pter 4
the modern biosphere— almost all the main animal phyla were
already present, and for the next 500 million years those same
players would organize themselves into elaborate oxygen-
dependent ecosystems with multitiered food webs, expand-
ing onto the continents and into the skies, developing ever
more specialized adaptations to their ambient environments.
And for the next 500 million years, they suffered spectacularly
whenever those environments, and especially the atmosphere,
changed too fast.
C U R TA I N S
In the nineteenth century, the field of geology was primarily
devoted to paleontology, and even before Darwin’s On the
Origin of Species was published in 1859, fossils were being used
to demarcate divisions of geologic time. Victorian geologists
studiously catalogued gradational changes in certain lineages
like the coiled ammonites, whose shells bear ornate patterns as
distinctive to particular moments in time as hoopskirts or saddle
shoes. But geologists also recognized points in the rock record
at which the changes in the fossils were not merely incremental
alterations in costume detail but wholesale replacement of one
cast of characters with an entirely new troupe. On the basis of
such discontinuities, John Phillips— the nephew of canal dig-
ger William Smith, who had developed the concept of index
fossils— proposed in 1841 that there had been three great chap-
ters in the history of life: the Paleozoic, Mesozoic, and Cenozoic
(Old, Middle, and Recent Life) eras. (The much deeper roots of
life in the Archean, more than 3 billion years before the start of
the Paleozoic, would not be appreciated for another century).
Phillips, an orphan, was raised by William Smith and ac-
companied him as a child on many fossil expeditions. He was
Changes in the air 117
an excellent and perceptive paleontologist but became a vocal
opponent of Darwin’s theory of evolution by natural selection,
believing instead that the exquisite match between animals
and their environments was evidence of a divine plan (which
apparently allowed do- overs). In the later part of his career,
Phillips allied himself with Lord Kelvin to undermine Darwin’s
assertions about the “prodigious durations of the geological
epochs.”20 Still, his chapter designations for the epic story of
animal evolution were astute.
Darwin was understandably irritated by Phillips but could
not deny that the fossil record did seem to have some sudden
and perplexing disappearances. Confident that evolution pro-
ceeded at a consistent pace, however, he did not perceive these
as evidence for natural catastrophes. Darwin fully accepted the
concept of extinction; indeed, the continual culling of organ-
isms was central to his theory. But he argued that what appear
in sedimentary rock sequences to be sudden extinction events
were simply artifacts of the intermittent nature of sedimenta-
tion. He devoted an entire chapter in Origin to “The Imper-
fection of the Geologic Record,” in which he emphasized that
rocks document only a fraction of elapsed time, stating, “be-
tween each successive formation, we have, in the opinion of
most geologists, enormously long blank periods.” Darwin also
suggested that the rates of sedimentation, when it does occur,
may not be fast enough to capture evolution in progress: “Al-
though each formation may mark a very long lapse of years,
each perhaps is short compared with the period requisite to
change one species into another.” He further speculated, per-
ceptively, that our reading of the fossil record is skewed by the
fact that we can find fossils only in settings where sediments
once accumulated (otherwise there is no rock), but those set-
tings are not always the places where organisms may have lived.
118 Ch a pter 4
Darwin’s inclination to explain away discontinuities in the fos-
sil record would prevail well into the mid- twentieth century,
when the geologic timescale was well enough calibrated to
make it undeniable that on occasion bad things had suddenly
happened to good ecosystems. We know now that there have
been at least five great mass extinctions, and many smaller ones,
since the start of Cambrian time. After each of these, life on
Earth eventually recovered but was irrevocably changed, with
the creatures that survived, as much by happenstance as hard-
earned fitness, becoming the unlikely founders of brave new
biospheres.
A P O C A LY P S E N O W
In a mass extinction, the normally meticulous scalpel of nat-
ural selection, which excises this moth or spares that finch on
account of the tiniest differences in wing color or beak shape,
becomes the evolutionary equivalent of a machete. Whole tax-
onomic groups of organisms— not merely individuals or species
but genera, families, and orders— in many locations and hab-
itats are cut down in swift, indiscriminate strokes. The cause
of a mass extinction is generally very different from the factors
behind ordinary thinning by natural selection, in the same way
that deaths from wars or epidemics differ in a fundamental way
from deaths due to individual accidents or illness. Paleontol-
ogists quantify the severity in terms of the magnitude of de-
viation from the background rate of extinction for different
groups. The background rate of extinction for amphibians in the
Cenozoic, for example, is less than 0.01 species/year or about
one frog or salamander per century.21 Mass extinctions imply
that the normally commensurate tempos of evolution and en-
vironmental change— well matched over time, in the same way
Changes in the air 119
that tectonics and erosion keep pace with each other— have
fallen out of synchrony. Gradual geologic change— the growth
of mountain belts, the separation of continents— inspires the
biosphere to innovate, but abrupt shifts may devastate it. In
mass extinctions, alterations to the environment have for some
reason accelerated to the point at which much of t
he biosphere
cannot keep up.
It is fascinating to look back at hypotheses for the end-
Cretaceous extinction described in the textbook for the Earth
History course I took in college in the early 1980s, just before
the Alvarez meteorite impact hypothesis began to gain traction
in the geologic community. Old evolutionarily untenable ideas
about the dinosaurs being sluggish and stupid— and by impli-
cation “deserving” of extinction— had by then given way to
new depictions of creatures that were sprightly, warm- blooded,
sociable (in some cases), and even smart. So killing them off
had become harder, and none of the proposals about their
demise— global cooling, virulent plagues, genocide by egg-
eating mammals, deadly allergies to the just- evolved flowering
plants (!)— seemed to be shocks sufficiently short and sharp to
do the job. The single extraterrestrial hypothesis mentioned
was the notion that cosmic radiation from a distant supernova
might have reached Earth just at the moment when there was
a magnetic field reversal and the planet was least protected— a
literal disaster in the Greek meaning of the word: “bad star.”
Reading these ideas now feels like revisiting a kinder, gentler
moment in history, because scientific ideas about mass extinc-
tions seem to parallel contemporary sources of existential angst
in society; the geologic past often acts as a screen onto which
we project our deepest fears. This is not to say that hypotheses
about mass extinctions are unscientific, but that terror of new
types of apocalypse helps fuel our imaginations about possible
120 Ch a pter 4
scenarios for cataclysms of the past. Geologists, as humans who
live in particular social settings and historical moments, cannot
help but be influenced by the prevailing zeitgeist. Compared
with the jittery angst of the twentieth and twenty- first centu-
ries, the Victorian period was a time of great optimism about
the potential of technological and scientific progress to improve
the lot of mankind. So besides the Lyellian taboo against in-
voking geologic catastrophes (specifically the old- fashioned
biblical type), it may be that because the Victorians were not
haunted by visions of the end times, Armageddon was simply
not in the scientific air.
In 1980, however, fearsome technological advances that
the Victorians could not have foreseen threatened human civ-
ilization, and it was at that anxious moment late in the Cold
War that the Alvarez meteorite impact hypothesis emerged.
Its description of a dusty shroud of pulverized rock blasted into
the stratosphere, blocking photosynthesis and leading to mass
starvation, came directly from the “nuclear winter” scenarios of
Carl Sagan and atmospheric chemist Paul Crutzen in the 1970s.
The eruption of Mount Saint Helens that same year made it
even easier to imagine an ashy doomsday.
By the time the Chicxulub crater was identified in 1990, the
Berlin Wall had fallen. As the threat of nuclear holocaust began
to fade from the collective consciousness, it was replaced by a
growing awareness that environmental malefactions might be
humanity’s downfall. Acid rain was shown to be devastating
forests in New England and Scandinavia, the legacy of sulfurous
emissions from decades of coal burning. The selective pattern
of marine extinction at the end of the Cretaceous, with shelled
creatures in deep water faring better than those in the shal-
lows, suddenly looked very much like what one would expect
in an ocean that had become soured by sulfuric acid. And the
Changes in the air 121
rocks in the Yucatán crater had plenty of sulfur in them: they
included thick layers of a mineral called anhydrite, or anhy-
drous calcium sulfate, which would have been vaporized in the
impact, hurled into the atmosphere, and then precipitated as
burning acid rain. The 1991 eruption of Mount Pinatubo, in
the Philippines— 10 times more powerful than that of Mount
Saint Helens— provided further insight. The eruption injected
enough sulfate particles into the stratosphere to counteract,
for two years, the inexorable climb in global temperatures re-
lated to rising greenhouse- gas concentrations. The immense
volumes of brimstone blasted from the 240- km (150 mi)- wide
Yucatán crater could have caused far more severe cooling—
devastating to organisms accustomed to the warm Cretaceous
world— before falling out of the atmosphere as the rain from
hell. It seemed, then, that sulfur, not just dust, must have been
the real culprit in the end- Cretaceous extinction.
But many paleontologists remained unsatisfied with this ex-
planation, too. Caustic acid rain should have been especially
harmful to freshwater ecosystems, yet species in these envi-
ronments, including frogs and other amphibians sensitive to
changes in water chemistry, had survival rates of close to 90%—
far higher than those that lived on dry land, where only 12%
withstood the cataclysm. The failure of any of the proposed
kill mechanisms to account for the details of the fossil record
has led some paleontologists to propose that the asteroid was
not a lone assassin but struck a global ecosystem already weak-
ened by other injuries. The most frequently cited accomplice
is volcanic activity, in particular the eruptions that produced
the Deccan Traps, a 1.6 km (1 mi)- thick stack of basalt flows
in present- day India. For tens of thousands of years leading up
to the extinction, the oozing lavas released enormous quan-
tities of carbon dioxide, creating a world that was already in
122 Ch a pter 4
environmental peril when it was mortally injured by a blow
from space. Vaporization of a thick sequence of limestone at
the Chicxulub site would have injected even more CO2 into the
air, so that after a few years of frigid cold from the pall of ash,
the climate whipsawed into a withering hothouse. In recent
reconstructions of the Cretaceous finale, the murderous but
charismatic asteroid has been forced to share the stage with far
less glamorous greenhouse gases.
B A D A I R D AY S
The study of mass extinctions became a distinct and fashion-
able subdiscipline within paleontology in the decade after the
end- Cretaceous impact was proposed. To those who embraced
the newly “legalized” catastrophism, it seemed likely that all
mass extinctions could eventually be blamed on extraterres-
trial impacts. A brilliant paleontologist Jack Sepkoski of the
University of Chicago, who was the first to recognize the po-
tential of Big Data in paleontology, believed he had detected a
26- million- year cycle in extinction frequency through an anal-
ysis of thousands of fossil catalogs. In a strange kind of neo-
uniformitarianism, he speculated that episodic die- offs might
be linked with Earth’s periodic passage through the spiral arms
of t
he galaxy, which could destabilize the orbits of comets.22
This inspired eager searches for evidence of large impacts at
the times of other mass extinctions, and moved the study of
impact cratering from a fringe field into the geologic main-
stream. But three decades later, no other major biological crisis
has been convincingly linked with the crash landing of a comet
or asteroid. We are left with the sobering fact that sometimes
things can go horribly wrong for life on this planet, for reasons
completely internal to the Earth system.
Changes in the air 123
Besides the end- Cretaceous cataclysm, the other great ex-
tinctions include, chronologically (1) the Late Ordovician
event about 440 million years ago, which was the first major
pruning following the Cambrian explosion; (2) a closely spaced
pair of die- offs in the late Devonian Period (about 365 million
years ago), by which time macroscopic life had moved onto
land; (3) the end- Permian holocaust 250 million years ago,
the mother of all mass extinctions, which John Phillips aptly
marked as the close of the Paleozoic Era; and (4) the Late Tri-
assic event, a cruel blow just 50 million years after the Permian
debacle. Depending on how one measures the severity of these
massacres (by numbers of species or genera or families van-
quished), the dinosaur extinction is the fourth or fifth in rank.
Although the victims and the circumstances of these calami-
ties differ in detail, they share some striking similarities (appen-
dix III). All— including the end- Cretaceous event— involved
abrupt climate change, and all, with the exception of the De-
vonian event (when tropical seas turned cold), are linked with
rapid warming. Second, all involved major perturbations to the
carbon cycle and carbon content of the atmosphere, either by
unusually effusive volcanism (Permian, Triassic, Cretaceous)
and/or through an imbalance between carbon sequestered by
the biosphere and carbon released from stored hydrocarbons
(Ordovician, Devonian, Permian, Triassic). Third, all entailed
rapid changes in ocean chemistry, including acidification that
devastated calcite- secreting organisms (Permian, Triassic,
Cretaceous) and/or widespread anoxia ( dead zones), which
asphyxiated almost everybody except for sulfur- loving bacte-
ria (Ordovician, Devonian, Permian). All the extinctions, in
Timefulness Page 14