Reconstructing prehistoric environments with pollen
With spring comes a wave of wind-blown pollen—the yellowish plant “dust” that has continually coated the landscape for millions of years. Though present-day pollen spells trouble for hay fever sufferers, prehistoric pollen is proving incredibly useful in reconstructing ancient environments.
Plants produce pollen to hold and protect the male genetic information required for reproduction.
Pollen travels through the environment by way of wind, water, or animals, bound for female flowers and cones. A lot of pollen grains never reach their destination. Under the right conditions, these “wasted” grains become part of the fossil record.
“We get essentially the ‘leftovers’ of reproduction,” said Dr. Surangi Punyasena, a scientist at the University of Illinois at Urbana-Champaign, who specializes in palynology, the study of pollen and spores.
“We don't know the exact process by which pollen goes from flower into fossil record,” Punyasena said. In general, though, a pollen grain must settle in an area with standing water, sink, and be buried by fine-grained sediment in the absence of oxygen to become a fossil, she said.
“The advantage pollen has is that it's actually designed to be preserved,” said Punyasena.
Pollen’s outer covering, or exine, is made up of a very resistant biomolecule called sporopollenin. “Sporopollenin is designed to resist UV and protect the very sensitive genetic material. You're essentially creating not an indestructible capsule, but a very resistant capsule already. That's the prime reason why pollen preserves,” Punyasena explained.
Unlike other types of fossils (for example, bones from vertebrates), the pollen record is both widespread geographically and relatively continuous through time. Further, it is representative of the surrounding prehistoric plant community per modern assessments.
Consequently, palynology can provide important insights into ancient environments and climates and the origin and extinction of plant groups that other fossil records cannot offer. The field has been used to study environments dating back to the origin of the earliest land plants, about 400 to 450 million years ago.
Unfortunately, palynology can be a challenging field to work in.
“It is not like the paleontology people think of, where you find the fossil when you’re out in the field,” Punyasena said. “You don't know what you have until you dissolve [the rock], and to do it you have use fairly toxic chemicals. You have to be prepared to do that. You have to be trained to do that,” she said.
After preparation, palynologists examine the tiny pollen grains with a microscope. Grains range from 4.5 to 200 micrometers in size in flowering plants—roughly the same width range as a human hair.
Palynologists begin with assessing the general shape of the pollen grain, then shift to the minute features of the grain’s exine.
“Pollen will have a huge diversity of ornamentation. This is one thing that makes identification complex, but it's also what makes the identification precise. Because you have the diversity of features, you can discriminate down to genus/species in some cases,” said Punyasena. “Diversity has a strong taxonomic component—things that are related look alike.”
A very detailed vocabulary is used to describe pollen features; however, identification is largely a visual skill.
“The analysts that work the fastest and the most precise essentially have this catalogue in their head of all the types they've ever seen,” said Punyasena. “It takes years, almost even a lifetime to recognize and learn these types. And then what happens is that people retire and die and that knowledge is gone. The next generation is restarting.”
There is no efficient way to pass on identification expertise currently.
Punyasena believes that artificial intelligence may hold the key.
Some neural nets, or computer systems that are designed to act like the human brain, are pre-trained to analyze images. Researchers can further “train” these systems by identifying and tagging features.
“In my ideal world, all palynologists are tagging—they're making their identifications virtually,” Punyasena explained.
Each new identification and tag added makes the system more efficient at recognizing and identifying pollen.
“Eventually, every pollen grain that has been identified by a palynologist resides somewhere. That way we're not wasting our time on identifying things that we can already identify. People are only working on the systems and the types that are new,” Punyasena said.
Editor’s notes:
To reach Surangi Punyasena, call 217-244-8049; email spunya1@illinois.edu.
