Pulses of extinction hit long-bodied and compact-bodied sea monsters simultaneously. The long-bodied eel-
shaped whale of the Eocene was just the last wave of evolutionary replacement that followed a mass extinction.
And the compact Eocene whales—like Zygorhiza—were the last wave of replacement of the fast-swimming guild.
Cretaceous sea monsters in
the deep and the shallows.
Marine habitats came in two
types: 1) the very shallow,
weed-choked seas that
spread north to south across
the middle of North
America; and 2 ) the clear,
deep water off the
continental coasts, like the
Pacific Ocean along
California. Long-bodied sea
lizards dominated the
shallows, but the deep
waters hosted much larger
populations of long-necked
swan lizards. Both habitats
were struck by mass
extinction at the end of the
Cretaceous.
vier's law of land-sea simultaneity applies. Each time the land eco-
system suffers mass extinctions, the oceanic system suffers as well.
As dinosaurs were snuffed out at the end of the Cretaceous, the
great sea lizards, and the snake-necked plesiosaurs were also dying
out, as were a host of large and small invertebrates, from coral-
like oysters to shelled squid and microscopic plankton. This same
land-sea simultaneity marked the Tartarian disaster, the Jurassic
extinctions, and all the other times of Great Dying, including the
one that struck our class Mammalia at the end of the Eocene Ep-
och during the Age of Mammals. The dual land-and-sea nature of
these extinctions automatically eliminates a long list of potential
agents. The Cretaceous die-off cannot be explained by the evolu-
tion of poisonous plants, for example, because the sea creatures
would not have been affected by their toxicity.
430 | DYNASTIC FRAILTY AND THE PULSES OF ANIMAL HISTORY
The next step is to observe that the extinctions hit the land
and the sea at the same time but in different ways. The entire salt-
water marine system suffered, the extinction being as complete
among the tiny, planktonic animals as among the giant sea ser-
pents. The extinctions eliminated most large animals but left
freshwater swimmers nearly untouched. Crocodiles, alligators, and
freshwater turtles changed little from the Cretaceous to the post-
Cretaceous. Another difference between the land and the sea was
that small land animals did not suffer as much as the large ones.
The big dinosaurs of the Cretaceous disappeared totally, but many
families of lizards and mammals passed right through the disaster
without losing their evolutionary stride. Finally—a point largely
ignored by most scientists—the land extinctions struck hardest at
the most dynamic, rapidly evolving groups of large creatures, the
families that showed the highest rate of producing species and the
most vigorous rates of adaptation. These are the very groups for
which there is the strongest evidence of warm-blooded metabo-
lism—the protomammals, the crimson crocodiles, the pterodac-
tyls, the dinosaurs, the large mammals of the Eocene Epoch.
Extinctions on land possess another peculiarity. On average,
plants do not suffer as much as the plant-eaters. The land plants
of the Late Cretaceous did suffer some extinctions but nothing
compared to the wholesale devastation experienced by the plant-
eating dinosaurs. The relative immunity of plants holds true for
the other previous and subsequent mass extinctions.
At this point, the modus operandi of the agent for the mass
extinctions is revealed in some detail. The suspect: (1) kills on land
and sea at the same time; (2) strikes hardest at large, fast-evolving
families on land; (3) hits small land animals less hard; (4) leaves
large cold-blooded animals untouched; (5) does not strike at
freshwater swimmers—most of these creatures are cold-blooded,
so criterion (4) applies; (6) strikes plant-eaters more severely than
plants.
Can any known agent of extinction operate according to this
pattern? At present many scientists do believe so. As of this writ-
ing, the scientific press is full of discussion about the newest so-
lution to the mass extinction—one that is literally unearthly. The
26 Million Year Death Star is supposedly a giant heavenly body
that strikes down the global ecosystems as it brushes past the earth
THE TWILIGHT OF THE DINOSAURS | 431
in a repeated series of near collisions. Such extraterrestrial theo-
ries are hardly new.
The impact of meteorites was proposed as the agent of ex-
tinction decades ago. No direct evidence for these existed until
Walter Alvarez, a chemist specializing in microanalysis of rare ele-
ments, discovered the now famous iridium layer. Iridium is a "no-
ble metal," similar to platinum but denser, and like platinum or
gold it does not easily form compounds with other elements. Ac-
cording to classic astrochemical theory, iridium is extremely rare
on the earth's surface, but much more abundant in celestial bodies
such as meteors, asteroids, and dead stars.
Alvarez did not originally start out to investigate for iridium.
His initial concern with Cretaceous sediment was quite routine;
he was trying to find ways of identifying the last layer of sediment
formed immediately before the Cretaceous Period ended. If that
layer could be identified, geologists could use that horizon as a
standard for comparing the sequence of geological events all over
the world. This type of marker has valuable applications, because
it is usually very hard to date strata in Europe relative to layers in
America. Occasionally layers of sediment carry distinctive chemi-
cal trademarks across millions of square miles because volcanic ac-
tivity can spread clouds of fine-particle dust over the world's oceans,
depositing unique concentrations of minerals in the mud at the sea
bottom.
Alvarez used detectors of ultra-high sensitivity on ocean sed-
iment at Gubbio, Italy, where an unusually good sequence of strata
was laid down at the very end of the Cretaceous. An abrupt change
in plankton fossils marked the layer where paleontologists would
place the end of the Cretaceous. Right exactly at that point Al-
varez stumbled upon a striking geochemical marker, with totally
unexpected implications—a thin zone rich in iridium.
Alvarez and his co-workers announced their discovery and
stunned the paleontological community with their conclusion: A
giant meteor had struck down the world of dinosaurs. The central
idea was that such a huge meteor (or asteroid), smashing into the
earth at the very end of the Cretaceous, would blanket an im-
mense area of the earth with its extraterrestrial cargo of iridium
because the explosion of the celestial mass would send up vast
clouds, full of iridium-rich dust. Such dust clouds would subse-
432 | DYNASTIC FRAILTY AND THE PULSES OF ANI
MAL HISTORY
Victims of mass extinction:
Long-bodied whales like
Basilosaurus and primitive
porpoiselike whales like
Zygorhiza died out at the
end of the Eocene Epoch.
quently settle down onto the earth's surface and meld onto the
face of the globe.
Walter Alvarez's theory gained converts essentially because a
giant celestial collision could explain part of the peculiar selectiv-
ity of the great extinctions. When the meteoric mass struck the
earth, the resulting dust clouds would blacken the sky, obstructing
sunlight. Plants would die, as would the plant-eaters and finally the
carnivores as temperatures dropped under the deadly umbrella of
dust. Cold-blooded creatures could hide in their burrows and wait
until the dust settled because their low metabolism would permit
long fasts with no ill effects. The lack of extinction among croco-
diles and turtles would therefore be easily explained. Small ani-
mals would suffer some extinctions (some mammal families did die
out), but they generally have burrows to hide in to avoid the con-
sequences of the dust-induced chill. And since many small animals
are omnivores, they could have survived by feeding on the car-
casses of the dinosaurs that had succumbed. Many species of plant
would survive in dormant states—seeds, spores, underground tu-
bers, and bulbs—so the mild effects on the plant world would also
be explicable. For those of us who are convinced the dinosaurs
were warm-blooded, the great dust cloud could explain why all of
them were wiped out. Their high metabolism combined with large
size made them especially vulnerable because they could not wait
out the disaster. Even the repeat nature of mass extinctions could
be explained. A comet could follow a regular cycle of crashes with
the earth, a trajectory of collisions repeated every time the com-
et's and the earth's orbits coincided. A mathematical analysis pub-
lished in 1983 claimed that such extinctions struck regularly, every
26 million years, so the agent has even been dubbed with a name,
the 26 Million Year Death Star.
I do believe that extinctions come in cycles. I do not believe
the theory of a bolt from the cosmos. An astronomer friend of
mine from Boulder challenged me about this. I advocate a wide
variety of heresies about the dinosaurs, so why could I not accept
the theory of their extinction based on the striking meteor and
the resulting iridium layer? My defense is simple. I champion her-
esies only if they fit the facts better than orthodoxy.
The theory of the great meteoric explosion fails to fit the facts
in one major area. It insists that the extinctions were sudden, cat-
434 | DYNASTIC FRAILTY AND THE PULSES OF ANIMAL HISTORY
astrophic. All the dinosaurs supposedly died out in a few dozen
years, or approximately that. But for quite a while now, orthodox
paleobotany has maintained the extinctions were spread over tens
of thousands of years or more. And no question, this time ortho-
doxy has got it right. Paleontologists working in Montana claim
they observe a gradual extinction of dinosaurs and Cretaceous
mammals and a gradual build-up of new groups of Mammalia, des-
tined for world domination in the next era. In fact, with few ex-
ceptions (Dale Russell of Canada's National Museum being chief
among them), paleontologists are in rare agreement. The last ex-
tinctions were not a single weekend of colossal slaughter but a
drawn-out process requiring thousands and even millions of years.
What we require here is a careful, bed-by-bed analysis of the
fossil faunas. From them a precise schedule of the extinctions must
be established. If the last dynasty of dinosaurs proves to have been
dwindling for millions of years before the iridium layer was formed,
the theory of the cosmic collision loses all its validity. Can such a
detailed timetable be worked out? Unfortunately, the easiest and
most popular way to do it is also the most misleading: Counting
the number of genera in the fossil sample we possess. It is impos-
sible to know all the genera of dinosaur that lived at one time. It
is only possible to identify a fraction of the total because many
were so rare in life they had little chance of becoming fossilized.
Thus the more skeletons we discover, the more genera we are likely
to find. The only reliable way to compare the quantity of genera
is to juxtapose formations that contain the same number of iden-
tifiable specimens. The earliest of the three Late Cretaceous for-
mations in Alberta, the Judith River, has produced several hundred
skeletons; yet the next layer up, the Scollard, has produced only
forty. Obviously there is an enormous drop in the number of ge-
nera between the two formations, but perhaps this is merely the
result of the smaller sample taken in the Scollard.
Dale Russell has used a mathematical procedure called "rar-
efication analysis" to correct the numbers here. Rarefication analy-
sis determines how many genera of dinosaur would be found in
each formation if the number of skeletons were the same for each.
He concluded that the big drop between the Judith and the Scol-
lard would not appear as large if the same number of specimens
were available from both formations. Russell therefore decided
THE TWILIGHT OF THE DINOSAURS I 435
there was no crisis present among the dinosaurs in Scollard times.
He insisted that no significant decrease in genera of dinosaurs oc-
curred until the very end of the last formation, the Edmonton-
Hell Creek, when they all died off at once.
Can a correct answer be found here? Did the extinctions be-
gin millions of years before the iridium layer was laid down, or did
they happen suddenly, precisely at the time of the alleged cosmic
collision? To answer this question it is important to remember that
both genera and species of dinosaur had been dying out all through
the Cretaceous—all through the Mesozoic, in fact. What made the
final Cretaceous extinctions special is that no new wave of species
appeared to replace those that had died out. In one sense, that is
the essential point of all mass extinctions—the rates of extinction
outpace the production of new species, so whole groups simply
run out of species entirely. But Russell is correct in arguing that
there must have been many more dinosaurs near the very end than
are known. New species and genera of dinosaurs undoubtedly kept
evolving until near the final days of the Edmonton—Hell Creek.
But he is wrong in insisting that the world of the dinosaurs was
suffering no ills before those last days. A very important ingredi-
ent of the ecosystem was already falling to dangerous levels back
in Scollard times.
The dangerously declining parameter of the ecology was
evenness, what ecologists call equability. When ecosystems are
healthy and well insulated from extinction, no single genus domi-
nates. There will be several nearly equa
lly abundant genera in each
ecological category—several large plant-eaters, several big meat-
eaters, and so on. Judith River times were precisely like that. No
single genus of dinosaur was dominant; the chief roles were shared.
Three large duckbills were common: Corythosaurus, Lambeosaurus,
Prosaurolophus, and the horned dinosaurs were represented by three
fairly common genera: Centrosaurus, Cbasmosaurus, and Styraco-
saurus. Such evenness is expressed in mathematical terms by what
is called "Simpson's index." The fauna of the Judith scores a 3.2
on the Simpson scale.
The Simpson index is formulated to respond to an implicit
question: What is the probability of meeting two individuals of the
same genus in a row if animals were being met at random? If the
fauna in a given area is very even, with many equally common
436 | DYNASTIC FRAILTY AND THE PULSES OF ANIMAL HISTORY
genera, two individuals of the same genus would not be met in a
row very often. Hence a high Simpson score indicates a low prob-
ability of two-in-a-row, and that means high evenness. Simpson's
index is easily computed: (1) Take the commonest species. (2) Take
its share of the total population, say one third of the total. (3) Square
this fraction and the result is the probability of meeting that spe-
cies twice in a row. Thus if a species represents one third of the
whole fauna, two-in-a-row probability is (V5)2, which equals V9. (4)
Now repeat this calculation for all species. (5) Add all the two-in-
a-row probabilities together, and divide into one to yield the in-
dex. For example, assume four genera of dinosaur made up Vs, lA,
V4, and Vo of the total. The two-in-a-row probabilities are V% V16,
V6, and V56. Converting fractions to decimals, the probabilities are
0.11, 0.06, 0.06, and 0.03. Adding together, they equal 0.26.
1-^0.26 is 3.8 units. And 3.8 is the Simpson index. By contrast,
if one species represented 99-99 percent of the fauna, then the
two-in-a-row probability would be 0.999, and Simpson's index
would be about 1.0.
The dinosaurs of the Judith River enjoyed a rich, even eco-
system, one of the most even ever evolved. But the next layer up,
the Scollard, is very uneven. One genus of duckbill, Saurolophus,
made up 75 percent of all the big specimens, and others were quite
Robert T Bakker Page 44