Dinosaurs Without Bones (32 page)

The gastrolith connection between non-avian feathered theropods with gastroliths in early birds was bolstered by the discovery of gastroliths in the Early Cretaceous bird
Yanornis martini
of China. These gastroliths, mostly composed of sand- and gravel-sized quartz particles, were in exactly the same place where the bird’s gizzard should have been. Another insight provided by this avian rock collection was how an absence of gastroliths in other specimens of
Y. martini
, but the presence of fish remains in one, implied that this species might have been switching its diet seasonally. Some modern shorebirds do the same, eating seeds and insects in the spring through fall, but chowing down on seafood during the winter. So this Early Cretaceous bird may have died before the winter while its gizzard was still full of gastroliths, which it would have needed to fully digest and process fibrous seeds and insects.

Other theropods with gastroliths include a few species of ornithomimids. In 1890, nearly a hundred years before the theropod–bird connection was firmly established, O. C. Marsh named one dinosaur
Ornithomimus velox
; bones of another species,
O. edmontonicus
, are the most common Late Cretaceous theropods in North America, and Asian ornithomimids are not exactly rare either. Because “ornithomimid” means “ostrich mimic,” it seems only appropriate now that some of these dinosaurs, like modern ostriches, also had gastroliths. Ornithomimids with gastroliths include the Early Cretaceous
Sinornithomimus
, the Late Cretaceous
Shenzhousaurus
, and a dozen specimens of an unnamed species, all from China. The most recent, and one of the most exciting examples of gastroliths in an ornithomimid, was reported in 2013. More than a thousand rocks, all relatively small, were in a skeleton of the Late Cretaceous
Deinocheirus mirificus
of Mongolia. The hundreds of gastroliths from the dozen unidentified ornithomimids ranged from sand- to gravel-sized and were angular to rounded, affirming how gastroliths are the most diverse and unpredictable of dinosaur trace fossils. All of this points toward how these so-called “ostrich mimics” were 70 million years ahead of ostriches in using gastroliths, meaning these modern birds are actually the “mimics,” not their non-avian predecessors.
These gastroliths, combined with their toothless condition, also imply that ornithomimids were herbivorous, lending to the still-radical concept of vegetarian theropods.

Given this knowledge both old and new about gastroliths in theropods, what about sauropods and gastroliths? The long-held assumption is that huge sauropods, many of which only had puny, pencil-like teeth, used gastroliths to grind their food. However, this idea is now seriously doubted. The biggest problem with the previous explanation is that these gastroliths, like those used for “buoyancy control” in marine reptiles, are too few to have made any real difference in digestion. For sauropods to have actual functional “gastric mills,” they would have needed many more gastroliths than the ones found in sauropod skeletons so far.

For instance, in modern ostriches and other birds that employ gastroliths to help with their food, these rocks make up about 1% of their total body mass. The Early Cretaceous theropod
Caudipteryx
matches this ratio, implying that these rocks served a similar purpose in its lifestyle. However, for sauropods, the proportion between gastroliths and estimated body mass was about 10% that of birds. So if an ostrich were scaled up to 50 tons (scary thought), then it would need about 500 kg (1,100 lbs) of gastroliths to digest its food, which is about the weight of the largest Harley-Davidson motorcycle. Yet the greatest mass of gastroliths described thus far from a sauropod (
Diplodocus
) was only 15 kg (33 lbs), which is about the weight of a Huffy bicycle. Also, gastroliths are relatively rare in sauropod skeletons, and if used for something as essential as food processing, these trace fossils should be much more common. This huge disparity between extant and extinct gastrolith-using animals led paleontologists to conclude that these stones surely were not used for the same purposes, effectively pulverizing the “gastric mill” hypothesis. Alternatives may not be so exciting, but include: accidental ingestion, especially if rocks were adhered to plant roots; mineral supplements, such as for calcium or trace elements; or used as separators for keeping fibrous food from bunching in their stomachs.

Nonetheless, the disproving of one explanation for gastroliths in sauropods doesn’t mean they suddenly vanished. For example, one specimen of the Late Jurassic sauropod
Diplodocus hallorum
(formerly named “
Seismosaurus hallorum
”) of northern New Mexico had more than 200 gastroliths, which were carefully mapped inside and around its former body cavity. These gastroliths were all igneous and metamorphic rocks with varying degrees of polish to them, and most were about 2 to 8 cm (<1–3 in) wide. An assemblage of 115 gastroliths, totaling about 7 kg (15.4 lbs), was also packed into a small area within a skeleton of the Early Cretaceous sauropod
Cedarosaurus
from Utah. Other sauropods with gastroliths directly associated with their bones include:
Apatosaurus
,
Barosaurus
, and
Camarasaurus
of the western U.S., as well as
Dinheirosaurus
of Portugal (all Late Jurassic sauropods), and the Early Cretaceous
Rebbachisaurus
from Morocco. In one instance, 14 gastroliths were directly associated with the remains of a juvenile
Camarasaurus
scavenged by
Allosaurus
, the latter indicated by tracks, toothmarks, and dislodged teeth. Several specimens of a Late Triassic prosauropod from South Africa,
Massospondylus
—the same dinosaur affiliated with nests, eggs, and babies mentioned in a previous chapter—had gastroliths too. Moreover, the skeleton of another prosauropod, the Early Jurassic
Ammosaurus
of North America, had gastroliths. In short, although gastroliths are relatively rare, having been found in less than 4% of all sauropod skeletons, these trace fossils are abundant enough in them that they should not be ignored, either.

So imagine you are a 20-ton sauropod and walking along a stream bank during the Late Jurassic. At some point during your stroll, you get a serious hankering for some geo-gastroliths, and you don’t know why, because your brain is smaller than that of a 0.075-ton (150 lbs) human. How do you pick the right rocks? The cerebral processing needed to discriminate between “good” rocks (solid silica-rich ones) and “bad” rocks (crumbly ones or limestone) can’t be too complicated. One thing is for sure, though: it can’t be based on smell, because most rocks lack good scents. It also can’t be from sound, although the distinctive crunching of silica-rich
metamorphic or igneous rocks underfoot might have given off tones different from those of, say, limestone or shale. How about touch? Your feet might have somehow felt the right rocks based on their size, shape, and surface texture, but that would have required an unexpected sort of sauropod sensitivity akin to the princess and the pea.

Hence you are probably left with sight and taste. Depending on your visual spectrum, brightly or darkly hued rocks of a certain size might catch your eye. If too small, though, these rocks might not be worth the effort. If too big, you risk choking to death, which would be very bad for transmitting your genes. So you needed proper search images for potential gastroliths before lowering your head to the ground, grasping a rock or two (or three) with your teeth, and gulping it down. Perhaps during the short time a rock was in your mouth, taste might have had some influence on your choice, too. Do taste buds in your mouth help say “Yes, swallow this!” and you comply? Or does the opposite happen, a spit-take of rejected rocks, striking and killing nearby small theropods like stray missiles?

Imaginative scenarios aside, sauropods, theropods, and other dinosaurs that ate rocks on purpose must have applied some sort of pattern recognition to select the ones best for them, for whatever reason. Based on the presence of gastroliths in the Early Jurassic prosauropod
Massospondylus
, this ability and its outwardly expressed behavior probably started in dinosaurs early on, such as in the Late Triassic Period. From an evolutionary view, one of the more interesting aspects of gastroliths is how prosauropods, sauropods, and theropods (birds, too) all share a common ancestor as saurischians.

In contrast, only a few ornithischians have gastroliths. These include the Late Cretaceous ankylosaur
Panoplosaurus
of Canada, the Early Cretaceous ceratopsian
Psittacosaurus
of China, the Late Cretaceous ornithopod
Gasparinisaura
of Argentina, and nearly no others. So although gastroliths are rare in the vast majority of dinosaurs, they are much more likely to be in saurischians, showing up in those dinosaurs from the Late Triassic through the Late Cretaceous, and then continuing in birds. That means this viewing, recognition, grabbing, and eating of stones has been going on in
dinosaurs for nearly 200 million years, and will keep on happening as long as birds are around and also need gastroliths.

Speaking of birds, another type of dinosaur behavior that was possible, but difficult to test scientifically, is that parent dinosaurs taught their young by example, showing them how to select rocks for their own internal collections. Mammals are of course well known for passing on skills to their offspring, such as how mother grizzlies teach their cubs to fish for salmon. But in the past few years, behavioral scientists are gaining a greater appreciation for how some species of birds, such as crows and ravens, learn through experimentation and watching one another, while also imparting newly acquired knowledge to their chicks. If any dinosaurs performed similar behaviors, such as teaching their offspring how to become rudimentary geologists, then this would be a glimpse of dinosaur learning.

One way for paleontologists to test this outlandish idea—that parent dinosaurs taught juvenile dinosaurs how to find the rocks they need—would be to look for gastrolith-bearing skeletons of the same species and at different growth stages. For instance, do both smaller skeletons (juveniles) and larger skeletons (adults) of the same species bear gastroliths, or do just the adults have them? If both have gastroliths, are these the same types of rocks, implying that they might have been picked out together and at the same places? What are the proportions of gastroliths to body sizes of the juveniles and adults: do they correlate, or are the numbers small enough in one of the size groups that accidental ingestion must be considered? These questions and more can be applied if gastroliths are some day found in multiple generations of the same species of dinosaur, allowing us to wonder beyond mechanistic explanations and consider the bendability of behavior.

Gastrolith Ghosts of the Triassic

Let’s say you’re a paleontologist and you’ve been asked to study a geological formation dated from the time when dinosaurs ruled the earth, or at least when they co-ruled it with insects. All of the rock types, sedimentary structures, invertebrate trace fossils, and
geochemical data in this formation show it was formed in environments that should have been inhabited by dinosaurs, such as river valleys, lakeshores, or seashores. Yet after weeks in the field spent scrutinizing these strata, you realize they not only lack dinosaur bones, but also are devoid of tracks, nests, eggs, toothmarks, coprolites, or other obvious evidence of dinosaurs.

The only possible sign of a former dinosaur presence are some annoyingly regular, rounded, polished quartz pebbles and cobbles, which are locally abundant in some sedimentary layers. With a great sigh, knowing that your colleagues will castigate you, you are forced to consider that these rocks might be gastroliths, the only clues that dinosaurs were present. With much resignation, you write “Gastroliths?” in your field notebook, draw a little sad face next to this entry, and wonder what to do next.

Along those lines, in an article published by Robert Weems, Michelle Culp, and Oliver Wings in 2007, they argued that the presence of many unusual, rounded, polished rocks in the Bull Run Formation of northeastern Virginia demonstrated that dinosaurs lived there about 200
mya
. They came to this conclusion despite a complete lack of dinosaur bones, tracks, nests, eggs, coprolites, and other fossil evidence of dinosaurs in the Bull Run Formation. This meant these gastroliths, which were also actual examples of exoliths, constituted the only signs that dinosaurs had been there and then. It was an ichnologically wonderful study, one that demonstrated how more than a hundred years of cumulative knowledge about dinosaur gastroliths can be applied to quite reasonably state “dinosaurs were here” on the basis of a pile of rocks.

The paucity of dinosaur fossils other than gastroliths in the Bull Run Formation was frustrating, because it was the right age (Late Triassic) for preserving evidence of early dinosaurs in North America. Unfortunately for dinosaur hunters who live east of the Mississippi River, however, this depressing situation is the norm. Despite more than 200 years of paleontological study, Mesozoic rocks of the eastern U.S. are embarrassingly shy about revealing dinosaur bones, although a few places—such as in Connecticut,
Maryland, Massachusetts, and Virginia—have plenty of dinosaur tracks. This general lack of dinosaur bones is credited to poor preservational conditions—not enough dinosaur bodies getting buried quickly and in anaerobic environments—and what may have happened to bones after burial, such as dissolution by acidic groundwater. Yet many of these environments may not have been all that good for preserving dinosaur tracks, either. For example, only a few theropod tracks have been found thus far in Late Triassic rocks of northeastern Virginia, despite these having been formed in valleys with abundant river floodplains and lakes. Finally, paleontologists and geologists in the eastern U.S. face a challenge that is not so overt in the arid western U.S. states: plants. This lush vegetation not only covers up or otherwise obscures rocks, but also destroys original bedding and other primary features, such as dinosaur tracks, via root activity and soil formation, often prompting eastern U.S. geologists to jokingly praise the virtues of defoliants.