From roses to dogwoods to rice, flowering plants are among the most diverse and successful organisms on the planet. More than 350,000 species strong, they’re beautiful, nourishing, and critical for their ecosystems. Yet how they evolved has befuddled evolutionary biologists since Charles Darwin. Now, thanks to cutting-edge technology and a chance find in the Inner Mongolian countryside, researchers have taken a big step toward understanding how flowering plants, or angiosperms, came to be.
“It’s like a mystery being solved on CSI,” says Douglas Soltis, a plant evolutionary biologist at the University of Florida who was not involved with the work. Tracing how the traits of these ancient plants led to the structures of flowers today, he says, is “pretty exciting.”
Angiosperms evolved about 125 million years ago and now dominate many of Earth’s landscapes. They reproduce via seeds, as do the evolutionarily older gymnosperms, which include pine trees, ginkgoes, and others. But angiosperms evolved some key innovations for seed production that likely enabled their success. Their seeds form in the carpel, a tubular structure that sticks up from the center of a flower and matures into a pod that holds seeds—peas or beans, for example—inside.
The carpel grabs the pollen and transfers it to a chamber called an ovary, where seeds develop. Angiosperm seeds are encased in an inner and outer layer; the outer layer helps form the hard coat of a pea or the colored surface of a bean, for example.
Angiosperms evolved from gymnosperms, but how carpels and the second seed coat arose has been a big mystery. Fossils have yielded a few clues, but “nobody has been able to show where carpels come from,” says Michael Donoghue, a plant evolutionary biologist at Yale University who was not involved with the work.
Over the past 8 years, paleobotanist Gongle Shi of the Chinese Academy of Sciences’s Nanjing Institute of Geology and Palaeontology and others have uncovered fossil gymnosperms with seeds covered with cup-shaped outer coats from the United Kingdom, China, and even Antarctica. They called the outer coat a cupule and proposed that it was the precursor to the outer coat, or integument, of angiosperm seeds. But no living plants have such cupules, and the fossils the researchers had found were of partially decayed plants, making thorough analysis impossible.
So Shi was intrigued in 2015, when a local from Inner Mongolia showed him a piece of rock with swamp plant fossils containing cupules with exquisitely preserved details. The rock came from a coal mining region known as Jarud Banner. Shi and Peter Crane, a paleobotanist at the Oak Spring Garden Foundation, scoured the area and in 2017, they finally pinpointed the mine that was the source of the fossil. Once a swamp that had been blanketed in volcanic ash, the rock from the coal mine was full of fossils preserved when silica from the ash infiltrated the plants’ cells.
The researchers sliced up the fossil-filled rock with diamond-blade saws and polished and etched the surfaces with acid to make “peels,” which they could examine under a microscope. “When we cut it,” we could see individual cells,” Crane says. They also conducted a computerized tomography scan of the cupules to determine their 3D structure (see video, below). The team found that, as with the outer seed coat in modern angiosperm seeds, the cupule tissue curved around the developing seeds.
After comparing their new finds with other fossils with cupules, the researchers concluded that all belonged to a group of plants characterized by cupules of different types, they report today in Nature. Their existence suggests not only where the second seed coat came from, but also how carpels came to be, as some of these cupules appear to have modified leaves that could have evolved into carpels.
“It is very exciting to see a new link forged in the evolutionary chain from early seed plants to angiosperm,” says Charles Gasser, an emeritus plant developmental biologist at the University of California, Davis, who was not involved with the research. “The overall path [to flowering plants] is now much clearer.”
What’s especially clear, Donoghue says, is that carpels might have arisen in this group of plants. Until now, researchers had focused on a fossil cupule plant called Caytonia, discovered in Yorkshire, U.K., as the closest relative to angiosperms. But Caytonia has no trace of carpels. Now, having a whole group of potential closest relatives with a variety of cupule structures, “gives us different ideas about where the carpel has come from,” Donoghue says. None of today’s conifers or other gymnosperms were around millions of years ago to give rise to angiosperms, Crane notes, and “which [cupule fossil] is most closely related to flowering plants is hard to tell.”
“What we need,” Donoghue says, “is more of these kinds of discoveries and more careful reconstructions.” But in the meantime, Soltis adds, “it’s supercool when we get something like this out of the fossil record.”