Animal corpses rarely defy the dictate of "ashes to ashes, dust to dust" to become fossils—and even if they do, they don't remain sturdy for long. By the time paleontologists get their hands on ancient remains, the fossils are incredibly fragile. So for decades, researchers have tended to do their close analyses with replicas instead. But in the past decade, 3-D printing has enabled a new solution: printing out copies of skulls and bones.
"My research is in the evolution of higher primates," says paleoanthropologist Eric Delson, who uses digital models and an Objet Eden260 printer housed at Lehman College in New York City to produce models of primates' skeletons. 3-D printing allows him to make accurate replicas without damaging the originals, and to generate larger versions of fossils and even reconstructions of lost bones. In fact, the digital models that 3-D printers use as templates make it possible to re-create the remains of long-lost ancestral primates, putting classical model-making methods to shame by printing calculated re-creations of bones that failed to form fossils.
Typically, paleontologists make models by covering an ancient skull or fragment with liquid rubber (latex or silicon) to create a mold. These molds, however, sometimes develop inaccurate bulges if the material clings too tightly to the fossil surface or bubbles if the sticky material dries outside of a vacuum environment. Making the molds is a time consuming process, as researchers and technicians must work slowly and carefully to avoid damaging fragile skeletons. Despite the labor, even a successful mold quickly becomes useless after producing about six to 12 replicas, or casts; over time, the wear and tear on it prevents casting of a high-fidelity match of the fossil's shape.
"Those casts are fine," Delson says, "but they require equally specialized people, equipment and care that you don't damage [the fossils]."
To reproduce a fossil using 3-D technology, on the other hand, you don't even need to touch the original skull. A CT or laser scanner can capture a skull's surface without covering it in goo. Laser scanners bounce beams of light off the skull, which reflect back to a sensor on the scanner, dividing the surface up into about a million points that the scanner can reconstruct into a surface. Computed tomography (CT) scans, on the other hand, substitute x-rays for light rays, dividing a shape up into cross sections without ever physically contacting it.
Because 3-D printing involves the use of a digital template, the procedure can create more than just replicas: It can blow up or shrink down the copies it reproduces. "The ability of the 3-D printer to modify size is one of the great things about it for both research and teaching purposes," Delson says. A small stone tablet etched with cuneiform letters might cause difficulties for researchers who want to read it and study the depth of the incised letters. But scanning the tablet produces a malleable digital incarnation that can be expanded to twice its original size to help researchers better understand how it was produced.
Digital models also make reconstruction easier. Take the remains of a famous pre–Homo sapiens hominin discovered in Ethiopia in the 1970s, Australopithecus afarensis, nicknamed Lucy: although dozens of Lucy's bones have been excavated, about 60 percent of the skeleton is missing. To model the complete fossil, researchers have had to reconstruct the missing pieces. If they unearthed the right hand but not the left, for instance, the right limb can be scanned, and the resulting digital template can be flipped into a mirror image, such that a plastic model of the left hand can be printed. The remains of other australopiths in the area also have become digital templates: the vertebrae of a larger individual might be the wrong size for Lucy's frame, but the digital scan can be shrunk to fit, then printed out and fit in place. Digital printing and 3-D models allow researchers to create more accurate replicas than traditional methods like sculpting would.
Three-dimensional printing makes it possible to reproduce australopith bones that no longer exist, having been lost to decay millennia ago. But what about the bones of species that paleontologists have never uncovered? The study of human evolution relies on the fossil record, but the vast majority of ancient skulls and bones faded into dust rather than becoming fossilized for posterity. Digital modeling can raise them—or at least their facsimiles—from the ashes.
Of course, digital modeling alone is not sufficient: Other tools come into play. Perhaps the most important is the use of an evolutionary tree, a type of "family tree" for species. The primate evolutionary tree stretches back about 65 million years—but how exactly did researchers reconstruct it? To understand an this tree, it helps to start with a simple family tree.
Children's DNA comes directly from their parents, and because siblings share the same DNA source, their genes are very similar. Their maternal grandparents contributed to their DNA through their mother, and to their cousin Bob's DNA via their aunt. Because they share a common DNA source with Bob, their DNA will also be similar—but not as similar as that shared by the siblings, because the DNA source that Bob and his cousins have in common is one generation further away than the DNA source that the siblings share. In general, the more generational (or geologic, in the reference of evolution) time that separates cousins from a common ancestor, the more differences will be seen in their DNA. The offspring of one of the siblings and Bob's child would be second cousins, one generation further removed from the common source of DNA, and so they would be even less closely related to each other.
Researchers can use this genetic information when they look at the DNA sequences of modern primates. For example, the DNA of humans and that of chimpanzees is more similar than that of humans and gorillas. Relatively speaking, if humans and chimps are first cousins, then gorillas are our second cousins. Applying the family tree insight, this distinction indicates that humans and chimps share a grandparent, whereas humans, chimps and gorillas have the same great-grandparent. The last common ancestor of humans and chimps lived more recently than the last common ancestor of humans, chimps and gorillas.
Comparing the DNA of different primates thus allows scientists to visualize the course of primate evolution. Orangutan DNA differs even more from human DNA, indicating that the last common ancestor of orangutans and humans lived even longer ago than the last common ancestor of humans and gorillas. As the DNA of more and more species is compared, a tree emerges. But DNA is not the only method of comparing species.
"Morphometrics" is the study of an organism’s form, and Delson specializes in a subfield called geometric morphometrics to describe his specimens. This method records certain "landmark" coordinates on a skull's surface, creating a three-dimensional frame that encodes information about the skull's shape and size. By comparing the 3-D frames of various species and specimens, Delson and his colleagues can quantify how closely related they are.
When a new primate fossil is uncovered, DNA testing is rarely possible: There is often no tissue available to test. Instead, dating methods that examine the decay of radioactive elements such as radiocarbon can reveal how long ago a specimen lived, and this time window, along with morphometrics—comparing the remains' shape with other fossils and known species—can help researchers arrive at its position on the evolutionary tree. But although these fossils fit some nodes of the evolutionary tree, others remain blank—the primates that were the direct ancestors of some branches died and decayed without entering the fossil record. When it comes to reconstructing the skulls of these ancestral primates, digital modeling rises to the fore.
Delson's geometric morphometrics can be used to infer and create a 3-D coordinate system for every skull on the tree. Based on these shapes, he, or rather, his computer program, can work backward to surmise what an ancestral skull looked like. He uses the evolutionary tree to determine how similar the older skull should be to modern primates' skulls: It will hew more to the shape of more closely related skulls, although more distant descendants can also contribute information about the skull's contours. Once the program determines which skulls will be used to find the final product, and how much importance each contributing skull carries, it can combine their morphometric frames to deduce the frame of the ancestor's skull—and from the frame, what the actual skull would have looked like. Finally, Delson can plug the digital reconstruction into the 3-D printer to produce a fossil replica of a skull that no longer exists.
How does a 3-D printer work? Read on: “3-D Printing Gets Ahead: How Does a Printer Make a Fossil?”