Fossil flowers prove to be a fertile field

Paleontologist discovers blossoming plant diversity in 100-million-year-old specimens.

When Sir Peter Crane began his career as a paleontologist in the early 1980s, he was one of few people who focused on fossil flowers. They were rarely preserved, or so went the conventional wisdom. But Crane and colleague Else Marie Friis learned to sniff them out.

“We turned up wonderfully rich and beautifully preserved flowers, which have illuminated the early history of flowers and flowering plants in a way that would’ve been previously unthinkable,” said Crane, the John and Marion Sullivan University Professor in Geophysical Sciences and the College.

Friis opened the first plot of new fertile ground with her discovery of an assemblage of 80-million-year-old fossil flowers in southern Sweden. Now Friis, Crane and others have turned up similar troves of fossil flowers all over the world, including Kazakhstan, central Europe, even Antarctica. “And the fossils can be quite abundant,” Crane said. “You can get literally thousands of flowers out of a good sample.”

Most of these specimens are tiny, no more than half an inch long but with exquisite detail. “They’re preserved as charcoal formed in ancient forest fires,” Crane said. “Charcoal is very resistant, and there’s lots of it in the fossil record. The problem is it’s rather brittle.” Consequently, well-preserved assemblages of fossil flowers were probably burned near a small channel or pond, into which they were gently washed, then entombed in sediments. The trick for paleontologists: find ancient sediments that have been shallowly buried.

“If they’ve been deeply buried, squashed, and subjected to pressure, then this kind of fossil material is not going to come through,” Crane said.

Millenia of diversified growth

Crane and Friis met as graduate students in the late 1970s and began collaborating in 1985, when they and the University of London’s Bill Chaloner co-edited a book on flowering plant evolution and its biological consequences; it was an outgrowth of a symposium on the topic.

In later joint research, Crane and Friis, now professor of Paleobotany at the Swedish Museum of Natural History, tracked down deposits of fossil flowers (angiosperms) in eastern North America and central Portugal. The rock samples they collected are about 120 million years old but no much more consolidated than modern-day mud.

In the laboratory they soak the sediments in water, sieve out the charcoal, clean it in acids, sort through it with an optical microscope, then pick out the most interesting-looking fossils for more detailed examination under a scanning electron microscope.

“The material is significant in providing rich assemblages of small, three-dimensionally preserved flowers from the earliest phases of angiosperm diversification,” Friis said. “Currently these two areas are the only places that have yielded this kind of fossil from rocks this old.”

The fossils enabled Crane and Friis, along with longtime collaborator Kaj Pedersen at the University of Aarhus in Denmark, to reconstruct the reproductive biology and evolutionary history of the earliest angiosperms. Combined with the pollen data from other locations, “the material shows very clearly that the first major differentiation of angiosperms took place over a relatively short time during the early Cretaceous,” she said.

When flowering plants first appeared approximately 120 million years ago, they were relatively rare and displayed limited diversity. “By the time you get to 80 million years, there’s almost nothing but flowering plants, and there’s a huge diversity,” Crane said. “And that diversity has continued to build ever since.”

The plant ‘CT-scan’

Details of the coverings around the seeds are important for differentiating the seeds of flowering plants from those other types of seed plants. The seeds of most seed plants have one covering around them, what botanists call an integument, Crane explained. But angiosperm seems generally have two such coverings.

“One of the key questions is, how did that come about?” he said. In 1985, Crane published a paper highlighting several non-angiosperm seed plants that seem to have two coverings.

The most optimistic interpretation of Crane’s data: “Here is a group of plants that got part of the way to being an angiosperm, but not all of the way,” he said. But Crane turned his attention to other projects, and his interpretation garnered little support.

Crane’s interpretation is now gaining vogue, partly because of renewed interest in the evolutionary history of plants stimulated by the application of molecular data, and partly because of a new research technique that Friis, Crane, and their colleagues have begun to exploit.

In a November 2007 Nature paper, Crane, Friis, and their colleagues described their application of synchrotron radiation to the study of plant fossils. The technique involves irradiating plant specimens with a strong beam of X-rays to produce a sort of CT-scan of the subject matter.

“This allows you to capture, and then visualize in three dimensions, all the internal structures of this charcoalified fossil material,” Crane said. The technique is exactly what Crane needed to verify his suspicion that certain non-angiosperm seed plants had developed double seed coverings. “We’re now completely certain that several key groups of seed plants have two seed coverings,” he said. “What remains to be seen is their significance in relation to those of flowering plants.”

“I don’t really care too much about how it works out,” Crane said. “But I do think we’ve now got something that really needs to be explained.”