Adult skeletons of dinosaurs are not that uncommon because their massive bones and durable enamel teeth stand a reasonable chance of becoming fossilized. Finding the fragile bones and skin of unhatched dinosaurs is much more rare, however, because their poorly calcified bones and delicate skin decompose rapidly after the embryo dies. At the time of our discovery, in fact, fossilized embryos were known from only a handful of different dinosaur species, despite hundreds of species having been discovered. The embryos of three kinds of meat-eating dinosaurs had been excavated from Cretaceous rocks of Mongolia, China, and Montana, including specimens of an oviraptorid, a therizinosaurid, and a troodontid, respectively. Less complete embryos of an unidentified species of meat-eating dinosaur had been collected from late Jurassic rocks of Portugal, and embryos from one kind of plant-eating duckbill, Hypacrosaurus, were known from Montana. But absolutely no fossils of embryonic dinosaur skin had ever been discovered before 1997. This poor representation of embryonic dinosaur remains explains why we were anxious to get our eggs prepared and see what was inside.
Fossil preparators are essential in any paleontological team. Even with the advent of modern techniques such as CAT scans, which allow paleontologists to peer inside fossil bones and skulls without manually
As with other hatchling dinosaurs, the skull of our embryos had a shorter snout and larger orbits than the adults. The reconstructed embryonic skull (bottom) is compared to what the skull of adult titanosaurs (top) may have looked like.
A reconstruction of the skull of a titan-osaur (top) is compared to those of other sauropods with pencil-like teeth, such as Diplodocus (middle) and Nemegtosaurus (bottom).
cleaning away the surrounding rock, most fossils still require some manual preparation before they can be studied. Usually the surrounding rock must be removed by hand with small tools, such as dental picks, needles, miniature sandblasters, and miniature vibrating tools, although sometimes the rock surrounding a fossil can be etched away in a bath of dilute acid. Preparing delicate fossils such as our embryos requires enormous skill and patience. Under a microscope, the fossil preparator must slowly and carefully pick away the rock surrounding the fragile bones with the mechanical tools of the trade, then protect the bones by applying thin coats of transparent glue. To prepare just one tiny fossil embryo can take days or even weeks of painstaking work, but this had to be done to reveal the clues to identify the kind of dinosaur that had laid the eggs.
In late December 1997 and early January 1998, Marilyn Fox, our preparator at Yale University, made an important discovery. Inside one of the eggs, she uncovered some minute skull bones and miniature teeth. We hoped that the shape of the skull bones and the teeth would give us clues to identify the victims in the eggs, but this is never easy. Because the animals were so young, the bony tissue in their delicate bones was not well developed, thus obscuring comparisons with adult sauropods, whose bones are fully formed. Furthermore, the embryonic bones were crushed against the lower eggshell as the fluid in the egg leaked out and the bones became fossilized. Unfortunately, this would make identifying these small fossils difficult.
Luis took the train up to Yale to examine what Marilyn had uncovered. He brought several specimens back to New York just before Rodolfo came up from Argentina in the spring to see the fossils and help us write the scientific paper to announce our discoveries. One of our eggs contained a nearly complete skull, an important clue for identifying the kind of dinosaur that had laid the eggs because the skull bones of different kinds of dinosaurs are usually quite distinctive. The skull bones in our embryo were similar in shape to those of sauropod dinosaurs, the long-necked giants that include diplodocids, dicraeosaurids, barchiosaurids, and titanosaurs. This tiny skull was also remarkable because few adult skulls of sauropods have been discovered, let alone skulls of embryos. In fact, only a handful of adult sauro-pod skulls had been found in South America. Several bones in the skull of our embryo were not preserved well enough to be identified,
Only two groups of sauropods have pencil-like teeth: Diplodocus and its kin (top two images) and titanosaurs (bottom right two images). Other sauropods have thicker and more spoonlike teeth (bottom left).
but those that were allowed us to make a fairly accurate reconstruction of what the skull had looked like.
As in any other baby, the skull of our sauropod embryo was large in proportion to the size of the body, even though the whole head was only about two inches long. Likewise, the eye socket in our embryos was probably slightly larger in relation to the rest of the skull than the eye socket in adult sauropods, another characteristic of most infant animals. In addition, a couple of embryos had tiny, pencil-shaped teeth. The crown, or upper surface, was formed by enamel, the same extremely durable material that forms the crowns of human teeth, as well as those of many other animals. Dinosaur teeth come in a variety of shapes: some are designed like steak knives to cut flesh; others form tightly packed assemblages that serve as a grinding surface for macerating tough leaves and other kinds of vegetation; others are less specialized and shaped like tiny leaves. Despite all this dental diversity, however, only two groups of dinosaurs have pencil-shaped teeth — like the teeth from our embryos — and both are sauropods. Pencil-shaped teeth evolved once in the common ancestor of dicraeosaurids and diplodocids, but similar pencil-shaped teeth also evolved within titanosaurs. Which kind of sauropod did our teeth represent?
Titanosaurs are especially difficult to place on the evolutionary tree of sauropods. Some paleontologists argue that they are most closely related to diplodocids and dicraeosaurids, believing that the pencil-shaped teeth that are typical of all of these dinosaurs evolved only once. However, most students of sauropods argue that titanosaurs are most closely related to brachiosaurids. These researchers suggest that titanosaurs and brachiosaurs inherited a small claw on the first finger from their common ancestor, and that this claw was completely lost in some later members of the group. Additional characteristics in the hip and hind limb of these dinosaurs support the idea that brachiosaurids and titanosaurs are closely related. According to this argument, pencil-shaped teeth evolved twice—once in the common ancestor of dicraeosaurids and diplodocids and again in the common ancestor of titanosaurs. The fact that certain sauropods thought to be primitive titanosaurs — animals whose skeletons are very much like those of typical titanosaurs —lack pencil-like teeth supports this latter interpretation of a double evolutionary origin for this peculiar type of sauropod dentition.
The embryos lived at a time in which titanosaurs were common, especially in South America. As mentioned earlier, we had found their skeletons weathering out of the rocky cliffs near Dona Dora's puesto. The fact that titanosaurs lived around Auca Mahuevo at the same time as the embryos, however, does not constitute adequate evidence to conclude that the embryos were titanosaurs.
Our embryos had pencil-shaped teeth, but could we find unequivocal evidence in the bones and teeth of the embryos that linked them exclusively to either titanosaurs or to diplodocids and dicraeosaurids? Most experts believe that diplodocids and dicraeosaurids died out long before the late Cretaceous, but there is no consensus on this point. Central to this debate is a late Cretaceous sauropod from Mongolia called Nemegtosaurus, whom some regard as a titanosaur and others regard as a survivor of the diplodocid-dicraeosaurid group that survived long after most of its relatives had gone extinct. When the teeth of the upper and lower jaws met as Nemegtosaurus chewed, the crowns of the teeth were abraded such that nearly vertical wear surfaces formed, as in at least some titanosaurs. Surprisingly, a few teeth of our embryos had similar wear surfaces on their crowns. Even though the embryos could not have been chewing food before they hatched, they were obviously grinding their teeth in the same way they would have done when they ate after they hatched. Perhaps they were just exercising their jaw muscles to prepare for life in the world outside. Thus far, we had established that the embryos, Nemegtosaurus, and pencil-like-toothed titanosaurs all shared the same kind of dental wear surfaces. So if Nemegtosaurus was a titanosaur, the dental evidence would suggest that the embryos were titanosaurs. However, the only known fossil of Nemegtosaurus is a skull, which is not enough of the skeleton to determine if Nemeg-tosaurus belongs with the titanosaurs or with the diplodocids and dicraeosaurids, because little is known about the shape of the bones in a titanosaur skull. Unfortunately, therefore, we had no basis for comparison. There was at least one late Cretaceous titanosaurid roaming the ancient Gobi named Quesitosaurus, but no skull of this animal has yet been found. So we do not know what its teeth looked like or whether the whole skeleton of this creature was like that of Nemegtosaurus. Thus, although we can say that the embryonic teeth from Auca Mahuevo belonged to sauropods, it is not clear whether the embryos were titanosaurids, diplodocids, or another of their close relatives.
Another of our eggs contained several leg bones that fit up against one another. Although they did not prove helpful in precisely identifying our embryos, they established the approximate size of the embryo inside the egg. The thighbone, or femur, was about four inches long, twice as long as the skull of the other embryo, indicating that this embryo would have been ten to twelve inches long when it hatched. In adult sauropods, such as Diplodocus, the femur is four or five times longer than the skull. Although we cannot be sure whether the embryos were closer relatives of titanosaurs or diplodocids, all of these dinosaurs were enormous, and it is clear that our embryos would have grown into some of the largest animals ever to walk on earth. But what about the suspected fossils of embryonic dinosaur skin? Did those show similarities to previously known fossils of adult sauropod skin?
Fossils of adult sauropod skin have been known since the mid-1800s. The pattern of skin ornamentation in our embryos looked similar to that of other sauropod dinosaurs, such as Diplodocus. The skin in sauropods is formed by polygonal tubercles of varying size that do not overlap with one another. Recent discoveries have shown that these late Jurassic Diplodocus had a row of narrow spines running along their tails, like those of crocodiles, and some scientists argue that this series of spines would also have extended along the back and neck. None of the patches of fossilized skin that our crew found indicate the presence of spines on our embryos, but we believe that the triple row of larger scales found on our babies did extend along the entire tail, back, and neck. The difference between this pattern of embryonic scales and the pattern of spines in adult Diplodocus and its kin might suggest that the spines were used as a visual signal that allowed individual Diplodocus to recognize one another, perhaps at the time of selecting a mate.
The skin of our embryos exhibited a diverse array of scale patterns. In one, as mentioned above, a triple row of larger scales crossed a field of smaller scales. In others, we found scales arranged in rosette patterns, in which a circle of eight smaller scales surrounded a large central scale. Still other specimens revealed several triangular scales that converged toward a central point, like the petals of a flower. But
The pattern of scales of the embryonic sauropods (top left) resembled the arrangement of scales of the adults (center and bottom left).
unfortunately, it is not possible to say exactly where these scale arrangements were located on the body because the patches of fossilized skin did not overlap identifiable bones in the skeleton.
Several titanosaur specimens, all apparently close relatives of the late Cretaceous titanosaur from northwestern Argentina called Saltasaurus, were known from the same late Cretaceous rocks that we were exploring near Auca Mahuida. Saltasaurus is well known because it is covered with a fully armored skin, presumably for protection from the large meat-eating dinosaurs that lived at the time. Although armor is common in other groups of dinosaurs, including the stegosaurs and the ankylosaurs, it is not usually preserved with the skeletons of sauropods. In fact, it was not until the 1970s, when Saltasaurus was first found, that paleontologists realized that some sauropods possessed a covering of bony armor. We noticed that the pattern of armor plating in the skin of Saltasaurus was remarkably similar to the pattern of bumps on the skin of the embryos from Auca Mahuevo, which made us wonder whether we would find bony tissue within the skin of our embryos.
To study the skin of our embryos more closely, we wanted to take photographs using a stereo electron microscope and cut cross sections through the skin to study under other microscopes. Our SEM at the American Museum of Natural History required the specimens to be coated with a thin layer of gold or platinum paint before images could be taken, but we did not want to coat our real specimens. (Ironically, they are much more valuable to us without a gold or platinum coating than with one.) So we produced a rubber mold of our most complete patch of skin and made a resin cast of the skin from it. This perfectly replicated patch of embryonic dinosaur skin was coated with a thin layer of gold and then photographed. Luckily, we also obtained images of our embryonic skin outside the American Museum of Natural History using a more sophisticated electron microscope, which did not require the specimens to be coated.
Even at high levels of magnification, our embryonic dinosaur skin looks quite similar to the skin of other reptiles, like a blanket of round, scalelike knobs of similar size. In contrast to the scales on most modern lizards and snakes, the scales of our embryos did not overlap one another, just as in the case of fossilized skin from adult dinosaurs. In this respect, the skin of dinosaurs, including that of our embryos,
Foot-long baby sauropods hatched from a nest laid 80 million years ago in a remote corner of Patagonia, Argentina.
looks more like the knobby skin of Gila monsters than that of typical lizards. Folds in the skin on our embryos indicated that the skin was not closely attached to the muscles and bones. The folds probably formed in the joint areas between bones, just as skin folds at joints in modern animals.
As we said earlier, the scale patterns on our embryos were similar to the clusters of bony plates called scutes that had been discovered around the skeleton of Saltasaurus. The armor of Saltasaurus is formed by hundreds of small, closely packed, bony scutes —roughly the size of our fingernails —which are occasionally separated by four-inch-long, oval scutes adorned with a central ridge. This combination of large and small scutes occasionally forms roselike patterns. Both kinds of scutes are thought to have "floated" in the dinosaur's hide, although on certain areas of the body the scutes are so tightly packed that they would have formed a pavement of armor. Since the discovery of Saltasaurus, diverse scutes of other titanosaurs have been found in other parts of South America, Madagascar, and Europe. These discoveries have prompted the reinterpretation of scutes from earlier discoveries of titanosaurids, which had led to the belief that other kinds of armored dinosaurs such as ankylosaurids lived in South America.
Cross sections of the embryonic skin patches did not reveal bone. Nonetheless, the striking resemblance between the patterns on our embryos' skin and on Saltasaurus made us think that the bony armor of the adult titanosaur could have been formed as a one-to-one replication of the embryonic pattern of scales. This one-to-one replication is typical of modern armored reptiles, such as crocodiles and Gila monsters. Once again, our discovery suggested that the processes that control the development of modern animals were at work during the growth cycle of ancient dinosaurs. Although we believe that the scales of our embryos might constitute the model over which armor formed in adults, the smooth surface of our embryos' scales did not show any of the central crests seen in the larger scutes of Saltasaurus and other titanosaurs. It may be that these crests represent overgrowths developed during the formation of the bony scutes, or that perhaps crests were not present on the scutes of all sauropods.
All in all, our forensic studies revealed that the embryonic bones, teeth, and skin from Auca Mahuevo belonged to a large group of
With a much smaller body than its relative
Argentinosaurus, the 30-foot-long titanosaur Saltasaurus had its body protected by an armor of large and small bony scutes.
dinosaurs called neosauropods, which includes famous dinosaurs such as Diplodocus, Camarasaurus, Brachiosaurus, Titanosaurus, and Argentinosaurus. In a sense, we had discovered the smallest fossils of the largest dinosaurs. Our trip had succeeded beyond our wildest dreams, and our crew's discoveries represented several firsts for paleontologists. Most important, our discovery represented the first indisputable embryos of sauropod dinosaurs. Even though thousands of eggs attributed to sauropods had previously been found in France, Spain, India, Argentina, China, and other parts of the world, no definitive embryos of sauropod dinosaurs had ever been found until our team's discovery. A few small fossils from young sauropods had previously been discovered, and some paleontologists had argued that they represented embryos; however, they were either too big to really be embryos, or they were not found inside an egg —the definitive proof that a specimen had not yet hatched. At last, we could be certain that at least some sauropods laid eggs and that the large eggs previously identified as belonging to sauropods actually were sauropod eggs. Our embryos were also the first dinosaur embryos ever discovered in the Southern Hemisphere. Most dinosaur nesting grounds are concentrated in the northern continents, so the discovery of a large, new nesting ground in South America contributed important new insights to our knowledge about the reproductive biology of dinosaurs from the southern continents. Finally, the eggs also contained the first embryonic dinosaur skin ever discovered. So, for the first time, we could sense what it would have felt like to touch an unhatched baby dinosaur that would have grown up to become one of the largest animals ever to walk the earth. The discovery of embryonic sauropod skin also led to other important inferences concerning the biology of dinosaurs. As noted earlier, some scientists have argued that Tyran-nosaurus was covered with feathers as a juvenile and that it lost those feathers as it grew into adulthood. This inference is based on the close evolutionary relationship between birds and Tyrannosaurus on the family tree of theropods. Sauropods are not theropods but are included with theropods in the larger group called saurischians. That adult sauropods and theropods more primitive than Tyrannosaurus and its coelurosaur relatives lacked feathers when adults was already known from several specimens with preserved portions of skin. Our discovery made clear that not all saurischians were covered with feathers during the early stages of their development, because our embryos were not.
Still, a lot more research needed to be done so that we could write and publish a scientific paper announcing our discoveries. As the preparation of the specimens continued throughout February, March, and April of 1997, we began to compile this research. Unlike articles in newspapers and magazines, scientific papers often take months or years to get published. The staff of the journal sent our paper would send it to other paleontologists for their comments and criticisms, a process called peer review, which helps authors improve the quality of their manuscript. Once other scientists make their
comments, the editors at the journal decide whether the paper is important and accurate enough to publish. Once a paper is accepted, it can take several months to well over a year before the journal can produce and publish the paper. But before submitting our paper, we needed to try to figure out when and how the embryos had died.
Was this article helpful?