Some land plants use silica, a hard, glass-like mineral, to grow taller, wider, and stronger. That silica is absorbed from the soil into plant leaves and stems and, when the plant dies, deposited into rivers that flow into the sea, where the dissolved plant silica is used to make shells for the tiny plankton that absorb roughly 50 percent of the carbon dioxide in the ocean.

Scientists had long believed that plant absorption of silica was a relatively recent phenomenon, geologically speaking. Conventional wisdom suggested that this process evolved within grasses that are the main reservoirs of plant silica today, meaning roughly 30 to 40 million years ago. New research by Jonathan Wilson, assistant professor of biology and environmental studies at Haverford, and colleagues proves, however, that this thinking was off by about 350 million years. 

"The silica cycle actually has a richer history," says Wilson, a paleobotanist, "because it’s not just from grasslands. It’s these other environments—tropical swamps from about 300 million years ago, where all our coal comes from—that don’t look anything like grasses, but probably had a large amount of silica in them. It changes the way you think about what the silica cycle might look like over geologic time."

Using a combination of bioinformatics and combustion experiments, Wilson and his collaborators set out to discover how long plants had been participating in this silica cycle. "Because," he says, "that helps us think about how plants might have been participating in the carbon cycle, in an unusual way, in deep time. If plants are fertilizing the ocean, and marine phytoplankton require silica, we want to know if there were times in the geologic past where plants were supplying more silica than today or less."

Wilson's team combed genetic databases of plants to identify molecular similarities between different types of plants that absorb silica. They discovered that four amino acids of similar chemical makeup were present in the silica-absorbers, forming tunnels in the pores of a plant's Casparian strip (an anatomical filter present in plant roots) that allow water to bring dissolved silica into the plant. Then the researchers set dried plants on fire to burn away the carbon and measure the silica that remained. 

Because it is a hard mineral that burns at incredibly high temperatures—think of glassblowing—silica leaves a sort of "skeleton" behind after the rest of the plant is burned away. Some of the plants that Wilson studied had as much as 10 percent of their total mass left behind in silica. "For plants that accumulate silica, when you put the remains under the scanning electron microscope, you can actually see anatomical structures of the plant," explains Wilson. "You can tell these parts of plant—xylem cells, the stomata of a leaf—are encrusted by silica as the plants actually grow."

In these experiments, all sorts of plants, not just grasses, left behind these "skeletons:" mosses, liverworts, ferns, rare seed plants, and many plants currently alive today that have a long fossil record. "We found this whole population of possible silica transporters from these ancient lineages that have probably been participating in the silica cycle for a long time—400 million years at least," says Wilson. 

The paper detailing the team's findings, "Four hundred million years of silica biomineralization in land plants," co-authored by Elizabeth Trembath-Reichert, Woodward Fischer, and Shawn McGlynn, appeared in the April 28 issue of Proceedings of the National Academy of Sciences (PNAS).

The next step of the researchers' work will be to try to understand why there has been such strong selective pressure for plants to evolve silica biomineralization. 

"It’s interesting to try to put this together," says Wilson. "One of the reasons I got into this line of work is that it’s kind of detective work. The data is lost, the story is lost, you end up with the murder weapon and the victim, and you have to work backwards—creatively—to put it all together. That might mean clipping some things from a botanical garden to bring back to the lab and set on fire, or it might mean learning how to solve the structures of proteins. It's following that question, wherever it goes. It's a lot of learning, and a lot of fun."

-Rebecca Raber

SEM image showing the silicified epidermis and vasculature of primitive land plant Selaginella courtesy of Elizabeth Trembath-Reichert.