FEATURE ARTICLE
Virtual Fossils from 425 Million-year-old Volcanic Ash
A set of exceptionally preserved but difficult-to-extract fossils reveals the diverse creatures from a Silurian sea-floor community
Derek E. Briggs, Derek J. Siveter, David J. Siveter, Mark D. Sutton
Liberating the Fossils
At the start, we studied these concretions by splitting them open with a hydraulic vise, cracking them in two and then further dividing the pieces in half until a fossil appeared—or until we reduced the sample to a pile of tiny fragments. About half of the concretions we examined in this way proved to contain fossils. But it was hard to glean much about the animals' complex shapes from these randomly split sections.
We did the best we could and attempted to discern the properties of the most common species, the tiny arthropod Offacolus kingi, by studying it in several hundred randomly split concretions. But despite our best efforts, the picture we obtained of this animal was woefully incomplete. We were unable, for example, to work out how its head appendages fit together. This approach was even less satisfactory for rarer species (which is to say, everything else)—just imagine trying to figure out what something like a shrimp looks like from just a few randomly oriented slices through it. So it became clear to us that we needed to find a way to extract the fossils from these rocks.
The calcite casts proved too small and delicate to be dug out physically, and they couldn't be dissolved out chemically because they are so similar in composition to the rest of the concretion. They were not visible in x-ray photographs or in the other scanning methods we tried because they have the same density as the rock that contains them. So we resorted to physical tomography, which is just a fancy term for serial grinding. That is, we ground away at the fossil in very fine increments, up to 50 per millimeter, and recorded each exposed surface as a digital image.
Paleontologists have used serial sectioning since the beginning of the last century. But working in the late 1990s, we were able to take advantage of modern computing techniques to produce high-fidelity visualizations of the data. That's not to say that the software for doing this was something we could buy off the shelf. One of us (Sutton) had to write a considerable amount of code from scratch. All this programming allowed a computer to distinguish automatically between the fossil and the lighter-colored rock enclosing it. We then edited these digital images to correct them where our eyes told us that such retouching was necessary. These two steps were the virtual analogues of what paleontologists normally do: digging bones out of rock and cleaning them.
It took a long time, but it was worth the effort. All our image gathering and editing produced spectacular results. Using Sutton's software, we can manipulate our virtual fossils on the computer screen, using stereo glasses to add depth. Or we can render them as rotating animations. What's more, the software allows different structures to be hidden at will, allowing us to perform virtual dissections of these long-dead creatures.
Revisiting our preliminary reconstruction of Offacolus, which was based on random sections, we were able to substitute direct observation for educated guesswork, allowing us to correct many minor errors. And we finally figured out the nature of the head appendages: There are seven pairs, five of them with two branches each. This better understanding allowed us to place Offacolus more accurately on life's evolutionary tree. This animal turns out to be a primitive member of the chelicerates, the major group that includes scorpions, mites, ticks and horseshoe "crabs."
Although our method of studying these fossils is time consuming and destructive, it has yielded a wealth of data unobtainable in any other way. Through these observations, the Herefordshire fossils have begun to give up their secrets, revealing a diverse and astonishingly well-preserved fauna.
» Post Comment