the gigantic Diplodocus, Brachiosaurus, and Brontosaurus.
III. The Age of the Low Feeders. The Cretaceous, when all the
terrestrial habitats were overrun by big beaked dinosaurs which
fed close to the ground. Each of these types had its own unique
approach to cropping the foliage, so each must have made a dis-
tinctive impact on the co-evolutionary history of plants.
Orthodox dinosaurology has muddied the conceptual waters
here by relegating the herbivorous dinosaurs to the swamps, where
they are supposed to have gummed nothing but water plants. Now
that we have a corrected view of them as dry-land herbivores, it is
possible to begin reconstructing a much more accurate context for
the Mesozoic evolution of the flowering plants. The first clue to
the interaction of dinosaurs and angiosperms can be found in the
timing of extinctions. Flowering plants first appeared in the Early
Cretaceous just after the extinctions which occurred at the end of
the second grand period (the age of stegosaurs and brontosaurs),
and as the replacements for the third grand period (the age of the
low feeders) were taking place. This sequence is highly suggestive.
When the coalition of stegosaurs and brontosaurs died out at the
end of the Jurassic, the plant-eating dinosaurs changed so pro-
foundly that the rules of co-evolution must have been reset. Could
this dramatic shift from Jurassic-type to Cretaceous-type dinosaurs
have opened the way for flowering plants? It's an exciting hy-
WHEN DINOSAURS INVENTED FLOWERS | 185
Dynasties of plants and plant-eaters
pothesis—that the revolution in dinosaur plant-eaters caused the
single most far-reaching development in the kingdom of land plants.
And the evidence suggesting such a cause-and-effect relationship
is very good.
To understand how the dinosaurs' plant eating changed and
how dinosaurs may have invented flowers, Stegosaurus and its strange
adaptations for eating plants must be understood. Here again or-
thodoxy has obfuscated some obvious truths about dinosaur bio-
mechanics. Stegosaurus is often portrayed as something of a misfit,
a quadruped endowed with two sets of mismatched legs—the front
pair too short and the hind too long. Stegosaur skeletons mounted
in museums pose the beast with a clumsy, shuffling gait, its hips
towering above its low-slung shoulders and its nose nearly at ground
level. Stegosaurus and its close kin were the only common large,
beaked dinosaurs in the Late Jurassic. Therefore its feeding habits
must have had a major influence on the evolution of plants. The
orthodox restorations depict Stegosaurus as an ungainly low crop-
per, plucking plants from within a few feet of the ground. That is
the precise inverse of the truth— Stegosaurus wasn't a badly de-
signed low feeder, it was a superbly designed high feeder.
The point the usual reconstructions miss completely is that
the plan of the stegosaur's body was carefully balanced so the beast
could rear up on its hind legs and tail to feed upright, in a tripodal
stance. The very tall bony spines at the hips become understand-
able only in that position. Tall vertebral spines supplied great le-
verage to the back's muscles and ligaments, so that the entire weight
of the trunk could be supported by the hindquarters. The brilliant
English biophysicist D'Arcy Thompson pointed out the mechani-
cal significance of tall spines at the hip as early as 1924. He dem-
onstrated that Stegosaurus's spine-and-ligament construction worked
much like a single-span suspension bridge. Such bridges are based
on a tall central tower (equivalent to the stegosaur's hind legs). The
span's vertical steel supports are shorter and shorter the further
they are from that tower for the sake of easier leverage (just as
the stegosaur's vertebral spines become shorter the further they
are fore and aft from the hips). Just as the thick cables of the sus-
pension bridge based on the central tower hoist the weight of the
span, so did the thick ligaments of the stegosaur's vertebral col-
umn based on the hips hoist the weight of its body. Modern cranes
WHEN DINOSAURS INVENTED FLOWERS I 187
Diracodon laticeps, a three-ton Late Jurassic stegosaur, tilting up its body with
a push of a hind foot
and derricks work on the same principle—the jib works like the
Stegosaur's backbone, the cables work like its back muscles.
All the details of Stegosaurus % construction meshed perfectly
to produce a body machine that could swing up easily from a four-
footed stance into a hind-legs-plus-tail posture. Its shortened front
legs reduced the dead weight forward, so the animal could hoist
its front end up with less effort. The base of the stegosaur's tail
was provided with huge bony flanges to anchor immensely strong
tail muscles designed to brace its body against the ground. The
stegosaur's ancestors had had stiff, brittle tails containing long bony
rods, like those of the duckbills, which held the tail and back rigid.
Clearly the stegosaurs had evolved away from the ancestral stiff-
tailedness and had acquired strong, supple tail joints all the way
from hip to tail tip, so the rear half of the tail could be flexed flush
against the ground. A masterful final touch completed the stego-
saur's tripodal equipment. In most dinosaurs, the lower bony prongs
at the tail's end (chevrons) were usually simple, straight spines. But
on stegosaurs those bones were expanded into a shape like the
runners on a sled, to provide a superior brace for the body's weight.
188 I THE HABITAT OF THE DINOSAURS
Tall vertebral spines give back muscles and
ligaments leverage for raising the body—just like a
crane's jib is braced by a cable. Stegosaurs had
much larger back-raising leverage than does an
elephant of the same weight.
Live stegosaurs might have been slow when moving on all
fours, but the excellence of their design would appear when they
walked into a grove of trees that appealed to their taste. With a
slight upward push from the front paws, the entire body would
pivot upward from the hips, raising the head up to twelve feet above
the ground—and now the stegosaur could poke its snout into high
shrubs and the low canopy of trees, selecting the choicest leaves
and branches.
As an undergraduate at Yale, I published several papers at-
tacking the orthodox theories and arguing for the tripodal habits
I have described here. I believed I had arrived at some quite new,
revolutionary ideas, until I discovered some papers actually pub-
lished in the last century. I had in fact merely resurrected a view
WHEN DINOSAURS INVENTED FLOWERS I 189
carefully worked out some eighty years earlier. Professor Marsh
had dug the first stegosaur skeleton in 1878, and many descrip-
tions published in the 1880s and 1890s portrayed the animals
rearing on their hind legs and tail, exactly as their bony anatomy
indicated. Somehow this totally logical interpretation was lost i
n
the 1920s; in its place orthodoxy substituted the nonsensical view
of stegosaurs as ill-designed quadrupeds.
Stegosaurus was not the only animal whose bodily configura-
tion was adaptively modeled for high feeding. The brontosaurian
dinosaurs, Diplodocus and Brontosaurus, evolved exactly the same
set of tripodal characteristics: short backs and short forelimbs to
lessen the weight of the trunk; tall, "suspension-bridge" vertebral
spines at the hips; huge and supple tail bones; gigantic muscles at
the base of the tail; and sledlike lower tail bones as final support
for the body's weight. When a Diplodocus raised itself into a tri-
podal posture, its feeding height was phenomenal by today's stan-
dards—its nose would have reached to forty feet or more above
the ground. Brontosaurus s
1
reach would have been a little less, but
Barosaurus, a close relative of Diplodocus, could have reached to
forty-five or fifty feet.
A modern giraffe at its full height can reach only up to eigh-
teen feet. Just as was the case with stegosaurs, these tripodal ad-
aptations of the diplodocines were clearly understood by the
pioneering American paleontologists; Elmer Riggs of the Chicago
Museum wrote a detailed explanation in 1904. But, as with the
stegosaurs, the orthodoxy of the 1920s forgot all about this work,
so that between 1930 and I960 the standard view likewise main-
tained that Diplodocus was a quadruped with maladaptedly short
front legs.
Every single one of the giant Late Jurassic dinosaurs was in
reality a high browser of some sort. Bracbiosaurus's tail was too
weak for a tripodal posture, but it compensated with its spectac-
ularly long neck, so that even on all fours it could reach up forty
feet. Camarasaurus was the shortest-necked brontosaur found at
Como. It could nevertheless stretch up to twenty-five feet with its
forelegs on the ground, and much higher if it assumed a tripodal
stance. Never before nor since the Late Jurassic has the world wit-
nessed such a profusion of high-feeding plant-eaters. This was
nothing less than a unique epoch in the history of herbivorous
190 | THE HABITAT OF THE DINOSAURS
Some plant-eating dinosaurs were designed for rearing up on hind legs and
tail. Brontosaurus and Stegosaurus had tall vertebral spines over the hips, so that the back muscles and ligaments had strong leverage for raising the body.
Other big herbivores had shorter vertebral spines and must have stayed on all
fours. (Each of these skeletons is drawn to the same length, hip to shoulder
socket, to show the proportions of the backbone for easy comparison.)
How dinosaur feeding style changed. At Como in the Late Jurassic, there were six long-necked brontosaurs plus a stegosaur. The stegosaur and three brontosaurs could rear up and feed tripodally (Barosaurus, Diplodocus, and Brontosaurus). But in Late Cretaceous habitats in Utah, there was only one kind of brontosaur— Alamosaurus—
and low-feeding beaked dinosaurs dominated the plant-eating role.
habits. No plant or leaf was safe from a dinosaur's mouth unless
it stood over fifty feet above the ground! Since no flowering plants
of any sort existed yet, the principal trees were broad-needle con-
ifers and tall, spiny-fronded cycadeoids. These plants had to be able
to protect at least some of their growth through a combination of
alkaloid poisons, heavy oils, or spiny branches. As conifers and cy-
cads grow slowly compared to many angiosperms, Late Jurassic trees
had to guard themselves carefully to obviate the destruction of en-
tire breeding stands.
During these Late Jurassic times, one danger was minimal.
Since few dinosaurs were specialized for feeding at ground level,
conifer seedlings and other sprouts wouldn't suffer the same de-
gree of cropping we find at work in the modern ecosystem, where
cattle, horses, sheep, goats, and many other mammals pluck their
plant food from the surface of the soil.
When the Late Jurassic curtain fell and the Early Cretaceous
began, most of the old, established high croppers died out. The
mysterious hand of worldwide disaster swept across the conti-
nents, killing off whole families of herbivorous species and thin-
ning the ranks of the surviving clans. Stegosaurs disappeared
forever. Hardly a species survived into the Cretaceous. Diplodo-
cus, Brontosaurus, and nearly all the other tripodal brontosaurs died
out too. Some of Brachiosaurus's relatives survived, but that was
an exception to the rule. This tremendous disaster destroyed the
high-browsing system permanently; it never recovered. And no
Cretaceous dinosaur evolved high-browsing adaptations to replace
Stegosaurus. A few new brontosaurs evolved— Alamosaurus, for in-
stance—but they never came close in number to the glorious days
of the Late Jurassic.
As the dust settled after these extinctions, the opportunists—
evolutionary carpetbaggers—started to move into the devastated
ecosystem. New herbivorous groups blossomed into clusters of new
species. And nearly all these new Cretaceous herbivores were
committed to cropping near ground level. Among the earliest of
the newcomers were the big, spike-shouldered nodosaurs. Igua-
nodons also evolved early in the Cretaceous, reaching weights of
two to three tons and developing broad muzzles for close crop-
ping. Parrot dinosaurs emerged shortly after, as did the queer-
looking domeheads, both groups with their backbones curved to
WHEN DINOSAURS INVENTED FLOWERS I 193
bring their snout close to the ground. More and more low feeders
appeared as the Cretaceous continued—the wide-snouted ankylo-
saurs, the large and small horned dinosaurs, the myriads of duck-
bills.
This was an unprecedented alteration in the nature of plant
eating. In place of the Late Jurassic's tall browsers, the Cretaceous
concentrated on munching close to the ground. Low shrubs and
seedlings now faced a threat from herbivores magnified many times
compared to the conditions prevalent during the Jurassic. In this
ecological context the very first flowering plants appeared on the
earth's surface. How did the Ur-angiosperms react to all this
munching close to the ground? One ideal adaptive strategy for a
194 | THE HABITAT OF THE DINOSAURS
low woody shrub or seedling would be to grow as fast as possible
to achieve a height where the low browsers could not threaten it.
Early angiosperms could do that—their basic reproductive equip-
ment gave them a fast-growing edge over many contemporaneous
plants. A second approach would be a "dicey" strategy—scatter
sufficient seeds onto an overgrazed patch of bare soil to produce
a clump of shrubs quickly, and thus spread another generation of
seeds before the herbivores returned and once again mowed
everything to the ground. Early angiosperms could do that too—
their seeds and flowers allowed them to spread more quickly than
conifers or other nonflowering plants.
The low-level cropping brigade of the Cr
etaceous. On a typical Late
Cretaceous meadow in Alberta, all the big plant-eaters were specialized for
feeding close to ground level. Broad-snouted duckbills bit off wide mouthfuls
of vegetation; delicate-snouted parksosaurs nipped precision bites of selected
leaves. And the other dinosaurs added their cropping activity to the medley
of ground-floor herbivory.
WHEN DINOSAURS INVENTED FLOWERS | 195
How Cretaceous dinosaurs invented flowers. Nodosaurs, iguanodonts, and
the other plant-eaters that fed near the ground threatened meadows with
overgrazing. Nonflowering plants (conifers, cycads, ferns) couldn't regenerate
as quickly as angiosperms and they couldn't recolonize as rapidly. So after a
heavy raid by the plant-eaters, flowering plants took over.
Obviously intense low cropping placed a premium on any and
all plant adaptations for fast spreading, fast growing, and fast re-
production. And early angiosperms performed exactly these bio-
logical functions especially well. From this, a quite plausible scenario
emerges: Low-feeding Cretaceous dinosaurs opened the way for
the initial waves of angiosperms. Conifers, cycadeoids, and other
non-angiosperms were probably far less adaptive for handling the
assaults of new Cretaceous herbivores. Anywhere the plant-eaters
196 | THE HABITAT OF THE DINOSAURS
thinned out the conifer groves and cycadeoid thickets, an oppor-
tunity for species of flowering plants to win a foothold in the hab-
itat was created. Early angiosperms were probably cropped just as
severely as their neighbors, but their basic adaptations permitted
them to continue growing and reproducing in the face of the in-
tense mowing action of the dinosaurs.
Footholds are crucial for major adaptive revolutions. Any new
group finds it difficult to break into an ecosystem already full of
old, established groups. Conifers were highly adaptive old-timers
in Early Cretaceous times—the conifers had begun indeed long
before the first dinosaur evolved in the Triassic. Cycadeoids were
old-timers too, already diverse in the Early Jurassic epoch. When
angiosperms first evolved, they were confronted with meadows,
woodlands, and forests full of long-lived, highly refined plants. Early
Robert T Bakker Page 20
