The fossil record has much to teach modern-day geneticists about structural development, says palaeontologist Geoff Watts.
Fossils and genes are icons from the opposite ends of the biological spectrum. At one extreme, there is palaeontology: low-tech, descriptive, reliant on fieldwork and still emitting a whiff of the 19th century. At the other, developmental genetics: high-tech, experimental, laboratory science, very much of the 21st century.
Or so you think, until you meet Per Ahlberg of London's Natural History Museum. To reach his room in the museum, you pass cupboards of fossils in the bits of the building that the public does not get into. In the room itself the only evidence of his principal research interest is a scatter of rocks emblazoned with the skeletal remains of lobe-finned fish from the Devonian period 390 million years ago. Of his other preoccupation - developmental genetics - there is no sign at all.
It was in Cambridge, ten years ago, that Ahlberg happened to hear a lecture on the developmental biology of the fruit fly. It convinced him that palaeontology was ready to take off in a new direction. People like him now spend their time creating family trees and demonstrating natural selection's happy knack of turning legs into paddles and forearms into wings. They line up the transitional forms dug out of geological strata, and chart the steps in the transformation.
Ahlberg believes that by studying how today's animals progress from embryo to adult, you can infer something of the biological mechanisms that underpinned evolutionary change in their ancestors. For him, developmental genetics and palaeontology are not quite as remote as they first appear. Modern genetic methods mean that this is quite different from the old debate on ontogeny (form) repeating phylogeny (origins).
"We do have something in common," he says. "We are both dealing with the emergence of form." All that differs is the time scale: a few days or months from egg to offspring; many tens of thousands of years for a new species.
Ahlberg argues: "Evolutionary change is not a set of transitions that successively transform adult to adult to adult down the generations. That's not how it works. The actual sequence is 'adult to egg, adult to egg, adult to egg'. The evolution of form proceeds through changes in developmental processes."
The small, but significant, differences between offspring and their parents, on which natural selection relies, arise through differences in the genes controlling development.
In practice, what does this mean? Describing his own research, Ahlberg explains: "The work I'm doing with Brian (Metscher, an American developmental biologist working with Ahlberg at the museum) involves looking at how the symmetrical tail fin of modern fish evolved from the asymmetrical, shark-like form of their ancestors.
"We can see this happening in the fossil record. Conveniently, there are also living examples of the before, during and after stages of this change."
These living examples include the paddle fish (a type of sturgeon that has retained the primitive form of the tail fin) and a couple of US freshwater species that have an intermediate tail shape. Frozen at a particular stage of evolutionary change, these creatures provide a living insight into the biology of their extinct relatives.
"We study the trajectories of their embryonic development, put them side by side with those of fish that do have symmetrical tail fins, and see where and how during development they begin to diverge," says Ahlberg. "It's then possible to infer a lot about how the repatterning that took place during evolution produced the tail we see in most modern fish."
Ahlberg chose this project not for its earth-shattering significance, but because it is "eminently do-able", a good model with which to prove the approach. If he is successful, it might even be possible to identify the individual genes that brought about this or that evolutionary innovation.
The thought of being able to say what a particular gene was doing 50 or 100 or 500 million years ago is an extraordinary prospect. But yet again Ahlberg is modest about the extent to which it represents a new kind of thinking in biology. "We already know that certain genes involved in forming body patterns are strongly conserved, and you can find them in organisms as far apart as fruit flies and humans. So these genes must have been present in their last common ancestor, probably more than 550 million years ago."
So, any biologist who knows that different species can have many genes in common will be familiar with this dizzying time perspective, stretching back into the pre-Cambrian era.
Ahlberg and Metscher recently organised the first conference designed to encourage palaeontologists and developmental geneticists to talk to each other. It was a sell-out, and of the delegates - a more or less equal mixture of both disciplines - only a few were defensive about their territory, says Metscher.
"Talking to people in both camps, I found a similar reaction," he says. "Many were interested in the idea, but some were still uncertain if it's practicable."
The key, he says, lies in collaboration, not colonisation. "Palaeontologists shouldn't try to become geneticists, or start learning how to do their own DNA hybridisation."
Ahlberg echoes that "Brian has recently been showing me nicely stained embryos of zebra fish and paddle fish. I look down the microscope and say, 'Oh, that's pretty!' But it's Brian who can say which colour shows which tissue. Likewise, I can bring out some mangled fish on a limestone slab, explain what it is and why its morphology is interesting, without Brian having to learn how to interpret squashed fossils."
For the future of this attempt at marrying past and present, Ahlberg would like to establish an interdisciplinary evolution and development group at the museum. "I think we have the potential here for generating a major new field - a new evolutionary biology of form."