This is a page from the Franklin & Marshall magazine archives.
Return to the current magazine site






Summer 2000
Limited Possibilities

by Marcy Dubroff
photos by Lori Stahl

Related Links



How many ways are there to make the world? Most modern evolutionary biologists, including Harvard's Stephen Jay Gould, will tell you that the process of natural selection can produce an almost infinite variety of organisms. Their premise is that so many things are possible that if life were to evolve again, the world would be completely different from the one we know. If you ask the same question of Roger D.K. Thomas, F&M's John Williamson Nevin Professor of Geosciences, the answer will be very different.


In his sunny lab on the ground floor of the Hackman Physical Sciences Laboratories, Thomas laughs as he tosses a handful of dice on a table. "Evolution is not altogether a game of chance," he says, fingering one of the ivory cubes. Thomas, a paleontologist and expert on evolution, believes that, "As in a chess game, there is an interestingly limited number of ways of doing things."

Thomas, along with his student collaborators, Rebecca Shearman '99 and Graham Stewart '95, recently won recognition for their research on this problem. Their article, "Evolutionary Exploitation of Design Options by the First Animals with Hard Skeletons," appeared in the May 19 issue of Science magazine, the world's largest-circulation general scientific publication and one of its most prestigious.

Thomas, a member of the F&M faculty since 1975, is a graduate of Imperial College in London, England. He earned his Ph.D. at Harvard, where Stephen Gould was one of his mentors. Thomas is the author of numerous articles, and edited the book A Cold Look at the Warm-Blooded Dinosaurs. In 1981, he earned F&M's Lindback Award for Distinguished Teaching.

His love for teaching brought him back to the classroom after a stint as associate dean for academic affairs at F&M from 1987-91. "I enjoyed administrative work, which presented intriguing new challenges," he recalls, "but I was drawn back to the classroom. I enjoy teaching in the widest sense of that word. Also, I realized that I would never get any of my research projects completed as long as I remained in the dean's office."


The project that resulted in Thomas' Science debut has its origins in an earlier study undertaken with his German colleague, Wolf-Ernst Reif, of the University of Tªbingen. In that study, they found that the outcomes of evolution are much more predictable than many biologists have supposed. "We showed analytically that all the good ways of building animal skeletons have been exploited," says Thomas.

That's not to say that individual species are predictable, Thomas emphasizes. "There are many different ways of putting the parts together. But our research shows that the basic designs of organisms are predictable. How many three-legged cats do you know? How many five-legged dogs? If there are complex organisms in any other part of the universe that need to mate and seek their food, I predict that they are bilaterally symmetrical."

This pattern of repeated evolution of similar structures is called "convergence," according to Thomas. "Consider the wings of bats, pterodactyls and hang-gliders," he says. "Each set of wings evolved independently, but they are all designed on the same principle, like a kite, with membranes stretched between rigid struts."

Expanding on this concept of convergence, Thomas and Reif recognized seven general properties of animal skeletons that define a set of possibilities they dubbed "the Skeleton Space."


"The Skeleton Space is a map of viable options for the design of animal skeletons," Thomas explains. Among the variables are: whether a skeleton is internal or external; composed of rigid material, like vertebrate bone, or flexible; composed of one element (a snail shell), or two (clam shells and jaws); and whether the skeleton is composed of rods, plates or solids.

Permutations of these variables produce 1,536 unique design elements, which represent, as Thomas likes to say, "a construction set we can use to build potential skeletons, just as a child creates structures and machines with Tinker Toys." However, that number is too large to analyze conveniently, and, according to Thomas, "it was difficult to represent all seven dimensions simultaneously."

So, Thomas and Reif whittled the problem down to 182 categories, a much more manageable number, defined by pairs of variables. This enabled them to compare the types of skeletons used by different groups of organisms. Their analysis showed that the good ways of doing things had been discovered by evolving animals, again and again.

Related Links

Roger D.K. Thomas homepage
Explore the data and documentation on which the Science article was based.

Stephen Jay Gould
Learn more about this heavyweight of evolutionary biology at this Stanford site.

Burgess Shale Fossils
See photos of fossils from this million-year old deposit.

Shapes of Nature
Take a peak at an F&M course, co-authored by Thomas, that explores the relationship between man and nature.


The next step was to find out how rapidly these design options were incorporated into animal skeletons. "It has been known for a long time that most major groups of animals appeared within the first half of the Cambrian period," Thomas says. "We decided to see how soon the different ways of building skeletons were exploited, as Nature is quick to take advantage of any possibility that is open to it."

Working closely, first with Stewart, then with Shearman, Thomas decided to concentrate on the 505 million-year old Burgess Shale, in the Canadian Rockies. Deposited in quiet water at the foot of a massive limestone platform, this rock unit is rich in fossils that document the earliest evolution of animals with tissues, skeletons and complex internal organs. What is unique about the Burgess Shale fossils, according to Thomas, is that their carcasses are unusually well preserved. They provide amazingly fine details of the structure of these early animals, allowing paleontologists to learn what these animals looked like and how they lived.

"The Burgess Shale has a remarkably broad representation of weird and wonderful organisms, as well as members of most major groups living today," Thomas enthuses.

Thomas points out that some of the organisms found in the Burgess Shale also occur in the Kinzers Formation in Lancaster County. Many of these local fossils are housed in the North Museum of Natural History. "The Burgess Shale, the Kinzers Formation, and rocks near Chengjiang, in China, are the only three areas in the world where several of these animals have been found," he says.

What Thomas, Shearman and Stewart discovered was that "within 15 million years of the first appearance of animals with skeletons in early Cambrian time, 146 of 182 possibilities had been exploited." That number represents 80 percent of the options presented by the Skeleton Space. This is consistent with other evidence in the fossil record "which indicates that the possibilities opened up by major evolutionary innovations are exploited very rapidly," says Thomas.

The trio was thrilled to learn, in March, that Science would publish their results. "We were excited about this," Thomas says, "as it will enable a broad range of scientists beyond the field of paleontology to learn about our work. It will be interesting to see what kinds of reactions we get from our colleagues. Some will probably be sympathetic to our interpretation of the data. Gould and others, who emphasize the roles of historical accident and shuffling of the genes in the course of evolutionary change, will probably be much more skeptical. So, we may get some lively debate."

Previous Story