contrast in bioengineering these two beaked clans displayed. The
Triceratops's snout was a mammoth set of pincers, with a sharply
edged upper and lower beak, narrow from side to side, and cov-
ered with horn. Such deep, powerful beaks must have given those
horned giants the power to slash and cut long, tough fronds and
branches—fodder probably too coarse even for the wide beak of
duckbills.
After he had completed his unorthodox treatment of duck-
bills, John Ostrom, in 1963, attacked the problem of Triceratops's
diet. To maximize their biting strength, jaw muscles require three
biomechanical properties: muscles must be thick, they must be long,
and they must have great leverage. Great leverage can be devel-
oped by designing the muscles so that their line of pulls is located
far from the jaw joint. John Ostrom showed that Triceratops and
its horned relatives evolved high bony cranks on their lower jaws
to move the line of muscle pull up and thereby increase leverage.
Triceratops chewing design
168
THE HABITAT OF THE DINOSAURS
The evolution of duckbills remodeled the lines of muscle pull in
the same way, but the horned dinosaurs went far beyond them when
it came to muscle thickness and length. All primitive dinosaur heads
had some basic design problems that caused difficulties when evo-
lution tried to enlarge the biting muscles. Most of the jaw muscles
in the skull were housed inside of bony compartments located be-
hind the eye sockets, so the outer bony walls of the skull tended
to limit possible muscle size. If the muscles grew too big, they would
bulge out during a strong bite with a force that would burst open
the skull bones—certainly maladaptive.
But Triceratops % ancestors required even larger jaw muscles
because those dinosaurs were locked into an evolutionary path
leading to diets of ever tougher, thicker foods. They needed an
escape from the limitations imposed by the architecture of the older
skulls—and they found one. On the top of the primitive dino-
saur's head, behind the eye socket, were a pair of holes, covered
in life by a tough membrane. Holes like these evolved many times,
probably because the stresses of chewing were concentrated along
a certain few trajectories in the head bones. The most effective
way to construct a head was thus to evolve thick bone where stresses
were great and holes were stresses were minimal. Large upper-rear
head holes (formally called "temporal fenestrae" in anatomical
parlance) gave the horned dinosaurs their escape route to freedom
for the design of their jaw muscles. As the head evolved, the rear
rim of the hole grew upward and backward, forming a gigantic frill.
Since the jaw muscle was attached to the membrane that covered
the rear rim, as the rim grew backward, so did the muscle, and the
entire mass of the jaw muscle could enlarge to unprecedented
proportions. Marks on the Triceratops's frill illustrate how far the
muscles had enlarged in both length and width: On a big skull, the
distance across the mass of muscles is often three feet and the
maximum tract of muscle fiber often three and a half feet in length.
These muscles must have delivered an astounding bite in life, with
a force greater by far than any other land herbivore in life's his-
tory.
Triceratops could use this prodigious biting power either for
nipping branches at the beak end or for cutting up the fodder into
smaller chunks with its teeth. Horned dinosaurs had cheek pouches
and could employ tongue—cheek coordination to keep chopping
DINOSAURS AT TABLE I 169
Jaw muscles in Psittacosaurus (above)
and Protoceratops (below)
the bolus into ever finer slices. The geometry of their teeth re-
sembled the duckbills'—several rows of teeth were packed tightly
together so that enameled ridges provided a self-sharpening cut-
ting mosaic. But the horned dinosaurs' teeth were arranged to
provide more of a vertical slicing action and less fine shredding
What did Triceratops eat? Ostrom suggested cycadeoid fronds,
probably a good guess. Cycadeoids were plants with large fronds,
their leaves resembling the cycads popular today in Florida as dec-
orator shrubs. Cycadeoids were so common that the Cretaceous is
known as the Age of Cycads. Both cycads and cycadeoids had fronds
two, three, even four feet long, characterized by especially strong
fibers and prickly pointed leaflets, so that cutting such leaves was
a nasty business. But a rich source of protein and calories lay in
those Cretaceous fronds, awaiting any beast that could evolve the
proper chewing armament. Horned dinosaurs were late arrivals;
they didn't make their evolutionary debut until halfway through
the Cretaceous. But once they got going, they developed with ex-
plosive success, proliferating species by the dozen, the result of
their mechanical prowess in chopping the previously inaccessible
fronds.
Beaked dinosaurs featured another adaptive device in their
plant-eating repertoire, an extra-long digestive tract for soaking and
fermenting stubborn plant tissue. Paleontologists usually dismiss
any theorizing about the soft parts of dinosaurs. Stomachs rot, in-
testines decay . . . both disappear without a trace in the fossil.
Ergo, all speculation about gastrointestinal tracts in the Dinosauria
is futile. This is a serious problem because obviously there's no
hope of understanding the dinosaurs' approaches to plant eating
without at least some knowledge of their innards. Some skepti-
cism about the study of digestive systems is justified—only very
rarely do sediments preserve direct evidence of inner architec-
ture. Digestive structures possess neither bones nor other hard
tissue, so the only way their outlines can be preserved is on rare
occasions when mud fills the stomach and intestines before the tis-
sue rots (a few Coal Age amphibians were preserved that way, but
no dinosaurs).
Christine Janis, a fellow graduate student at Harvard in the
mid-seventies, was the first to excite my interest in the digestive
DINOSAURS AT TABLE | 171
organs. She pointed out that teeth and jaws tell only half the story.
Nearly all plant-eaters have fermenting vats, enlarged chambers
where food sits and soaks while microbes attack it with powerful
enzymes. Janis stressed how enormous is the variation in location
chosen by natural selection for the fermenting site. Ruminants—
the deer—cattle—antelope family—chose a forward site and re-
modeled their stomach into a complex multi-chambered rumen
where the bolus is soaked by enzymes. Since the rumen is located
in the forward stomach compartment, a deer, antelope, or buffalo
can crop leaves, wad them up into a bolus, pass it down for pre-
softening, then pass it back up to the teeth for a thorough chew
after the leaves have been softened. Forward locations offer sub-
stantial advantages—teeth are saved from unnecessary wear when
all food is pre-soaked and softened. Horses, rhinos, and ele-
phants, on the other hand, chose a rearward location, a pocket
evolved far back in the intestine or colon. Rearward location has
one major disadvantage—the bolus can't make any sort of return
to the mouth. But since the rear of the body cavity is spacious, its
advantage is that rearward fermenting vats can be huge.
A dinosaur's fossilized ribcage can reveal a great deal about
the organs it housed—information largely ignored until recently.
For one thing, how big the dinosaur's digestive chambers were can
be gauged by the size of the ribcage. Orthodoxy maintains that
many dinosaurs were too weak-toothed to eat tough plants; but
large digestive tracts could compensate for weak teeth. Janis made
a point often ignored by bone-and-teeth paleontologists: The bet-
ter the enzyme soak given to food, the fewer the teeth needed to
deal with any specific food texture. A good case in point: Today's
herbivorous lizards usually have relatively small, weak teeth and
until recently had the reputation of being inefficient plant-eaters,
but recent experiments show that some lizards carry out very ef-
fective rearward fermentation in their extra-long intestines. Giant
ground birds—rheas and ostriches—have tiny heads and no teeth
whatever, yet these birds successfully employ rearward fermenta-
tion on a large scale.
The precise details of any dinosaur's plumbing cannot be de-
termined, but the overall body contours outlined by the ribcage
and hips do show how large the entire digestive apparatus was and
in what locations. Humans don't have ribs in their abdominal sec-
172 | THE HABITAT OF THE DINOSAURS
tor—their ribs end at the posterior edge of the lung compartment.
But dinosaurs had long ribs attached to every segment of the
backbone from chest to hip, so the cross section of the digestive
tract is preserved by the skeletal architecture. Brontosaurs clearly
had short, deep, crowded gastric tracts, because the abdominal
ribcage was compact front to back, and the ribs over the belly arched
widely outward from the backbone like barrel staves. In general
configuration the brontosaur's intestines followed the proportions
of modern elephants. And, just as in elephants, the front edge of
their hip bone (ilium) was flared outward to support their wide
belly. Elephants are big rearward fermenters, and brontosaurs must
have been so too. Some brontosaurs had larger digestive organs
than others; Brontosaurus itself had a very short torso from front
to rear and must have had a less voluminous intestinal apparatus
than Brachiosaurus with its long torso. Equipped with both gizzard
stones and rearward fermenting vats, brontosaurs could have tackled
really tough plant food.
Early beaked dinosaurs were bipeds, using hind legs alone for
their fast locomotion. That presented a special design problem: They
had to evolve a large digestive tract without upsetting the balance
necessary for bipedal walking. In most primitive dinosaurs, the
digestive tract ended where it butted against the wide pubic bones,
which formed a rear bulkhead for the entire abdominal cavity.
Located above the pubic bones was a narrow passage through the
hips, through which must have passed all the animal's outlets—
colon, urinary tube, and birth canal. This pubic bulkhead was an ob-
stacle to redesigning the intestines: in order to lengthen the gastro-
intestinal tract, the pubic bulkhead had to be pushed backward and
with it the entire pelvis. So a longer digestive system implied a
longer, heavier body cavity before the hips—a shape hard to
balance. Primitive beaked dinosaurs were forced to face this design
problem squarely from the very beginning of their evolution-
ary development, because nearly all the earliest species were very
long-limbed and lively, built for ultrafast bipedal running. Fast lo-
comotion placed special strain on the back, and a long, heavy
stomach in front of the hips would be difficult to support. Evolu-
tion was thus pulling in two opposed directions—shorter torsos
provided better balance for running, but longer torsos were re-
quired by the need to deal with the problem of digesting leaves.
DINOSAURS AT TABLE | 173
How shifting the pubis backward lengthens
the guts in beaked dinosaurs. Arrow
points to pubis.
Escape from this biomechanical dilemma was engineered by a
clever bit of anatomical sleight-of-hand. Peter Galton was the first
to work out the details in his Ph.D. thesis in 1966. If beaked di-
nosaurs managed to shift their digestive system beneath the hip
bones, its length could be increased without relocating the hip joint.
That effect was obtained by bending the pubic bones backward from
the point where they already attached to the other hip bones, so
that the intestines could lengthen while the hip joint remained as
it was. In their new position, the pubic bones slanted downward
and backward instead of straight down as before. Beaked dino-
saurs finally shifted the lower end of the pubic bones so far back
that they were far behind the hip socket. In this remodeled posi-
tion the thick coils of intestines and the colon continued without
interruption backward from the belly, to below the hip socket, and
all the way to the base of the tail.
174 | THE HABITAT OF THE DINOSAURS
The solution was elegant. The digestive system could be
lengthened and bipedal balance improved at the same time, for
more intestinal tubing located to the rear of the hip joint clearly
helped to balance the weight of the body before it. Although the
beaked dinosaur classes are totally extinct nowadays, a quite vivid
picture of their intestinal arrangements can be obtained from a
chicken, turkey, or any modern bird. In birds, the gastrointestinal
tract passes right under the hips all the way to the end of the pubis
below the base of the tail. And birds obtain the same advantages
for balance and strength as did the dinosaurs. Birds—like early
beaked dinosaurs—are bipeds when they walk, and their intestinal
arrangement allows them easy balance on their hindlegs. In the air,
all the force of the wingbeat passes through the upper shoulder
joint, and the design of the intestines lets birds have very short,
very strongly braced torsos to anchor the stresses and strains of
flying.
Some beaked dinosaurs had every possible plant-digesting
device—the parrot dinosaurs, for example, had (1) strong, deep
beaks, (2) closely packed teeth, (3) large jaw muscles, (4) a fore-
stomach gastric mill with stomach stones, and (5) a long intestine.
These parrot dinosaurs were the ultimate in dietary adaptation. They
were, however, exceptions to the rule. Most advanced beaked di-
nosaurs heavily developed one type of digestive device or an-
other, not all at once. Duckbills possessed cranial Cuisinarts par
&
nbsp; excellence. They had developed the best teeth for shredding leaves
and twigs into tiny bits. Large horned dinosaurs were the long-
slicers, the best at cutting tough vegetables into digestible chunks.
Yet neither duckbills nor horned dinosaurs have ever been found
with gastroliths, even though dozens of good skeletons have been
hewn out of the Late Cretaceous rocks. Probably these strong-jawed
dinosaurs substituted the power of their teeth for stone power in
the stomach. And the size of their digestive system varied consid-
erably in these species: wide-beaked duckbills had broad tail bones
that supported enlarged colons and intestinal appendices; hollow-
crested duckbills and horned dinosaurs were less enlarged in the
rump.
The successful development of rearward fermenting vats pro-
vides the solution to one of the biggest puzzles about plant-eating
dinosaurs—the weak-gummed giants. Although brontosaurs,
DINOSAURS AT TABLE | 175
duckbills, and horned dinosaurs were all apparently well provided
with grinders of one anatomical sort or another (teeth or gizzard
stones or both), two groups of big plant-eating, beaked dinosaurs
seem to have been totally unprepared for grinding plants: the ar-
mored ankylosaurs and the dome-headed dinosaurs. Both groups
seem to have followed the wrong evolutionary path—their teeth
are smaller, weaker, and less tightly packed than in the ancestral
early beaked dinosaurs. Ankylosaur and domehead had tiny crowns
on their teeth, and the teeth were loosely spaced in the huge bony
jaws. Traditional paleontologists looked at those mouths and con-
cluded they couldn't have chewed anything tough with teeth like
those. These dinosaurs must have had restricted diets of soft food
and therefore low metabolic rates. As usual with such interpreta-
tions, they immediately arouse suspicion. After all, ankylosaurs and
domeheads were advanced dinosaurs bristling with specialized neck,
skull, limbs, vertebrae, and armor. So the very logical question is:
Were there any equally advanced gastrointestinal adaptations that
might have compensated for this chewing apparatus? There was
indeed: giant afterburners.
No American museum presently displays an entire ankylo-
saur or domehead, even though a dozen good skeletons lie in