consisted entirely of shredded leaves.
Moas, of course, were birds, and birds are brontosaur nieces
(descendants of brontosaur relatives). The moas therefore present
an unparalleled opportunity to study how evolution equipped a
long-necked, pinheaded giant plant-eater with a high avian metab-
olism. And how did moas keep their body furnaces stoked with
sufficient fodder? Gizzards. Moa adaptations are known in great
detail, because specimens have been found with eggs (some con-
taining unhatched moa chicks), with pieces of skin, feathers, foot-
prints, and gizzard linings. Stones found within moa skeletons often
represent types of rock found nowhere else in the entombing sed-
iment. Moa bone-and-gizzard sites have been surveyed for miles
around in the search for the original source of the gizzard stones.
132 | THE HABITAT OF THE DINOSAURS
It is often found that the birds must have traveled as much as ten
miles to acquire pebbles of the desired consistency. Such careful
selection of gizzard stones implies that the moas were driven to
seek the hardest rocks for their gizzards, even if long searches were
required. Why expend so much effort looking for hard pebbles?
The high polish of moa gizzard stones provides the explanation.
Unlike turkey and alligator gizzard stones, which are often pitted
and dull, moa stones glisten with a fine patina which can be achieved
only by constant rubbing in very hard grit. We must infer that the
high polish was acquired in the course of day-to-day gastric func-
tion. The constant grinding of the hard pebbles proves that the
moa's mill had to operate nearly continuously to pulverize the great
masses of leaves and twigs needed to meet its metabolic needs.
Moas prove that a pinheaded brontosaur could process enough
fodder to support a high metabolism // the dinosaur possessed a
similar powerful food mill, equipped with very hard rocks. Can
unequivocal evidence for dinosaur gizzards be found? Yes. A few
years ago in the Victoria Museum in Rhodesia (now Zimbabwe),
its curator, Mike Reath, showed me dinosaur treasures excavated
from the red Forest Sandstone of the Zambesi Valley. One splen-
did skeleton belonged to a small, long-necked pinheaded dino-
saur, Massospondylus. Nestled within its ribcage was a neat cluster
of rounded, finely polished pebbles of a rock quite unknown else-
where in the sandstone. The curator related how a careful search
had demonstrated that such pebbles could not be found any closer
to the skeleton's quarry than fifteen miles. The conclusion he drew
was the same as the New Zealanders had drawn for the moas. This
very primitive Zimbabwean dinosaur was finicky about the quality
of its gizzard stones and would go out of its way to find only the
most resistant pieces. Massospondylus is a critical case in point for
the study of gizzard functions because it is closely related to the
direct ancestors of the great brontosaurs. If Massospondylus, a
brontosaur uncle on the dinosaur family tree, was equipped with
a gizzard like a moa, it wouldn't be surprising to find Brontosaurus
itself so equipped.
Perfectly preserved gizzard linings are perplexingly rare as
fossils. The Zimbabwean Massospondylus is unique, to my knowl-
edge, in preserving the gizzard lining from this species. Many other
specimens have been excavated from the forest sandstone, but none
GIZZARD STONES AND BRONTOSAUR MENUS | 133
show gizzards so well preserved. Why did the gastric rock disap-
pear before fossilization? Probably the gizzard rocks fell out of some
carcasses as they were ripped apart by scavengers. Other dinosaur
bodies probably became bloated after death, floated awhile in
floodwaters, then dropped their gizzard contents when the stom-
ach burst from the gases of decomposition. Some dinosaurs may
have belched up their gizzard lining while in their death throes.
Whatever the causes, it's unusual to find good fossil gizzards even
in those species that certainly had them in life. A good example
of how rare fossil gizzards are is provided by crocodiles and alli-
gators. Although crocodilians of all sorts possess muscular, rock-
lined gizzards, only a tiny fraction of the hundreds of good fossil
croc specimens preserve the gastric mill in recognizable form. So
the rarity of fossil gizzards in more recent species is worth re-
membering when the gastric functions of the more ancient dino-
saurs are discussed. Even if dozens of skeletons without gizzard
stones are found, it is not certain the species in question lacked
stones in life.
I don't know of a single brontosaur skeleton that shows a per-
fect pile of polished gizzard pebbles. However, I have seen a half-
dozen brontosaur bodies in the field where smoothly rounded
pebbles were scattered through and around the ribs. Could these
be the gizzard's contents, a bit displaced after death? I'm firmly
convinced they are. In each of these six cases the skeleton lay on
an ancient floodplain surface and the bones had been buried by
fine-grained sediment. In each case large, polished pebbles could
be found only near the brontosaur bones, nowhere else for
hundreds of yards in every direction. Streams can and do polish
hard pebbles to the same high patina found on the pebbles around
the brontosaur skeletons. But the geological circumstances sur-
rounding these six specimens absolutely rule out action by any
stream. The size of the grains of sediment is too fine to indicate
anything but the gentle slosh of mud-rich water flowing over the
floodplain. There is no way water could have moved those pol-
ished stones across the plain to the brontosaur sites. If water didn't
move the pebbles, then they must have traveled to the site inside
the dinosaur's stomach, to be deposited when the great beast
breathed its last and collapsed on the fern-covered meadow.
Other experienced dinosaur diggers have told me that they
134 | THE HABITAT OF THE DINOSAURS
too have found irregular patches of polished stones littering bron-
tosaur gravesites. And there is a further clue: the phantom stom-
achs. Years ago, when the pioneering brontosaur hunters rode on
horseback across badlands etched in the Morrison Formation, they
noticed piles of polished rocks lying isolated on the outcrops, no-
where near any dinosaur skeleton. I've seen them too, not only in
the Jurassic strata at Como but in the overlying sediment layers
deposited during the early days of the Cretaceous. Old-timers and
rock hounds sometimes call these polished-pebble heaps "dino-
saur belches." Chickens sometimes cough up their gizzard lining
when it gets worn, so that they can restock with fresh rocks. So
maybe the twenty-ton brontosaurs did the same. A seventy-footer
might feel a bit out of sorts. It would stop feeding, an involuntary
convulsion would ripple through its gizzard and forestomach, then
through the long neck, until out from the Brontosaurus's scaly lips
would drop a b
ushelful of outworn rocks. If such a scenario of di-
nobelches has any truth to it, geologists would have to take the
rock-transporting function of brontosaurs very seriously. Four or
five brontosaur species coexisted at any one place during this pe-
riod, together with two multi-ton species of stegosaur. Such num-
bers imply that rock-carrying gizzards were potentially quite
abundant all over the landscape. If each big dinoherbivore were
equipped with an outsized rock grinder, and they all regurgitated
a couple of times each year, then the Morrison landscapes would
have been the passive recipient of an endless series of pebble
showers from belching Stegosaurus, Brontosaurus, Camarasaurus, and
many others.
I was naturally skeptical the first time I heard this belch-a-
bushel theory from a wizened old Utah rock hound at a shop in
the Eden Valley of Wyoming. After all, he also was selling fossil
algae as "dinopoop." However, the phenomenon of patches of
pebble found without any related bones demands an explanation.
These masses of alien pebbles had to be dumped on the flood-
plains by some agency to be subsequently covered by a blanket of
fine mud. There is no evidence that the current of water carrying
the mud blanket was strong enough to roll these big polished peb-
bles over the meadows. Furthermore, no known hydraulic
mechanism could concentrate the pebbles in heaps. Dinobelches?
Could be.
GIZZARD STONES AND BRONTOSAUR MENUS | 135
An alternative theory derives from the experience of paleon-
tologists in New Zealand. Diggers of moa skeletons found not only
cases of gizzard stones within the ribcages, but also stone piles in
"ghost" skeletons—bones almost totally destroyed by soil water and
erosion. For their gizzards, moas preferred types of rock that were
highly resistant to geochemical decay. Pebbles rich in quartz suf-
fer little harm though buried for millennia in stagnant, soggy soil.
But bones will soften as the bone mineral (calcium phosphate) dis-
solves and the bone's connective tissue rots. New Zealand geolo-
gists found perplexing piles of pebbles scattered over the
countryside—pebbles too big and too concentrated in location to
be the product of any natural processes. And sometimes these
pebbles are foreigners, rounded fragments with a crystalline com-
position completely out of place where the piles of pebbles are
found. Only some unusual agent could have transported these
masses of pebbles. I am convinced some of the New Zealand peb-
ble masses are either moa belches or phantom moa stomachs—
from carcasses where bones have rotted leaving only the gizzard
stones. Almost certainly therefore some of the Jurassic and Early
Cretaceous piles of pebbles are ghost dinosaur stomachs.
The certainty of gizzards in dinosaurs has been with us for a
long time. A tiny skeleton complete with gizzard stones in perfect
order was found in Mongolia and was announced in popular and
scientific publications in the 1920s. Gizzards in dinosaurs would
discredit the time-honored orthodoxy which preaches that bron-
tosaurs had to be sluggish. We may even entertain the notion of
warm-blooded brontosaurs as a viable possibility.
By themselves, brontosaur gizzards don't indicate how much
or what these dinosaurs ate each day; other lines of evidence must
be employed to explore these questions. But brontosaur gizzards
and teeth together indicate what brontosaurs did not eat. They
didn't eat soft, mushy vegetation. Birds that subsist entirely on soft
fruits don't possess muscular gizzards and don't use hard pebbles
for their gizzard linings. Soft, watery food requires only a short,
simply constructed gut—with just enough contractile force to
squeeze out all the juices.
Brontosaur teeth, moreover, confirm the heretical idea that
they ate a tough vegetable diet. If the brontosaurs dined only on
soft water plants, then very little wear would be found on their
teeth. But in fact the teeth of Camarasaurus, Brachiosaurus, and
136 I THE HABITAT OF THE DINOSAURS
their kin manifest very severe wear, which could only have been
produced by tough or gritty food. Like the dental battery of other
dinosaurs, the teeth of these brontosaurs were continuously re-
newed. As one tooth wore out, it was pushed out of its socket by
a new tooth growing from beneath. So the wear on a single Cam-
arasaurus tooth represents the abrasion not from an entire bron-
tosaur lifetime but from a much shorter period of use—perhaps a
year or less.
Most shed Camarasaurus teeth are scalloped out on their front
and back edges by wear against some tough food. Such wear is
especially impressive considering the large size of the tooth's edge.
Camarasaurus teeth are very big—up to an inch and a half across.
Each camarasaur front tooth is something like a thick wooden
spatula in shape and is coated by a thick and roughened layer of
enamel. What sort of food could wear the broad grooves in such
teeth? Twigs and branches with tough bark, or big, palmlike fronds
from the cycadeoid trees which flourished all through the Jurassic
Period.
Diplodocus among the
conifer needles. Conifers
dominated the forest
canopies all through the
Jurassic. There were
conifers with tight-packed,
pineconelike needles
(Brachyphyllum—upper left),
and spirally arranged sharp
needles (Pagiopbyllum—
right), and very long,
pointed spear-shaped
needles {Podozamites—lower
right and in the Diplodocus
mouth).
GIZZARD STONES AND BRONTOSAUR MENUS
137
Camarasaurus and its kin represent the thick-toothed bron-
tosaurs. The slender-legged Diplodocus, on the other hand, was a
member of the pencil-toothed family. It was the Diplodocus % mouth
that inspired much of the orthodox view of pinheaded, weak-
toothed, and therefore sluggish brontosaurs. Compared to Camar-
asaurus, Diplodocus really does seem poorly supplied with dental
equipment. Diplodocus's teeth are limited to the very front of its
jaws; there are no teeth whatever in the posterior position where
most mammals have molars. Each Diplodocus tooth is very thin, and
all the teeth in the row are packed closely together in an arrange-
ment much like a miniature log palisade from an old-time West-
ern fort. Diplodocus did, almost certainly, employ gizzard stones.
In the field I have seen two specimens together with scattered
polished rocks. However, Diplodocus must have been a much more
careful eater than the Camarasaurus. Its teeth do, however, show
severe wear; usually the tips of the crowns are beveled from
grinding against some resistant food items. So at least its food wasn't
simply soft leaves and mush.
Consideration of the Diplodocus is complicated by a published
account of the "stomach contents" of one specimen. Alleg
edly, this
carcass contained the remains of a last supper preserved in the area
of the forestomach. The menu was strange: clamshells, bits of wood,
and bone fragments. Perhaps, as the authors of the published re-
port argued, Diplodocus was a scavenger picking through leftover
hunks of meat, shellfish, and whatever else.
I have learned that this "junk-food Diplodocus" is a hoax. Utah
geologists who know firsthand of the discovery have informed me
that the skeleton in question was a badly shattered brontosaur
preserved in a stream-bed sandstone. The alleged last supper was
not found within an undisturbed ribcage but was located very gen-
erally in the area of the torso. So the bits of clam and bone were
most likely placed there by the regular process of stream currents.
Patterns of tooth-wear in another Diplodocus specimen sug-
gested to one scientist that the pencil-toothed brontosaurs were
specialized clam-eaters. When the upper and lower jaws were closed
in this Diplodocus's skull, the worn tips of its upper teeth didn't fit
against the worn tips of the lower. To wear the teeth in this fash-
ion, the Diplodocus must have been biting wedged-shaped objects,
and since clamshells are wedged in shape, it is not totally impos-
138 | THE HABITAT OF THE DINOSAURS
sible that a clam-nibbling habit caused the wear. But there are many
other possibilities. If the Diplodocus used its teeth to strip leaves
from conifer branches, for example, then the gritty bark could
equally have beveled the tips of the teeth.
An important clue to the feeding habits of the pencil-toothed
brontosaurs is to be found in their peculiar head—neck posture. In
most living reptiles the neck is more or less horizontal, and the
head stretches straight forward from the end of the neck. Birds,
on the other hand, have a posture like our own—the head is held
horizontally but the neck is vertical and joins to the back of the
skull from beneath. In birds and people, the neck joints must ac-
commodate this erect posture, so the joint surfaces of the skull
face downward, not backward, as they do in lizards and crocodil-
ians. Diplodocus's head—neck joint was very much like ours: the
back of the skull faced downward, so the joint with the neck per-
mitted the snout to be horizontal when the neck was held upright.
The thick-toothed brontosaurs also had a deflected skull—to—neck
Robert T Bakker Page 14