A picture of drought

Nate Swenson strides so quickly through the Wisconsin forest while carrying a large pole clipper that postdoctoral researcher Vanessa Rubio usually follows the 40 feet of rope dragging behind him.

When they reach the designated plot, Swenson extends the clipper about 30 feet high and pulls the rope to snip off a leafy twig from the canopy of a tall, tagged tree. The twig floats down through the dappled sunlight and lands in his hand.

It would be a majestic scene but for the mosquitoes. Hundreds and thousands of them, swarming everything that smells like warm blood.

Swenson cuts twigs from nine sample trees in each plot where the reflection of light from the leaves could show up in a space-based image. Rubio selects one leaf from each twig, folds it into a labeled test tube, and drops it in a metal canister of liquid nitrogen.

This flash freezing will preserve its RNA, which degrades quickly otherwise. Later in his campus lab, Swenson can study the leaf’s gene expression, which changes as it is stressed by drought later in the summer.

Another leaf from the same twig is removed at the lab at the University of Notre Dame Environmental Research Center (UNDERC) near Land O’Lakes on the border of Wisconsin and Michigan’s Upper Peninsula. Rubio and Swenson shine a field spectrometer on the leaf to get a computer reading of its spectral reflectance and reveal its chemical and structural content, which likewise change due to drought stress.

It’s a painstaking process that will be repeated over hundreds of tagged trees at UNDERC in June and again in August. For trees too tall for the pole clipper, students shoot a line over a branch with an air cannon. The goal is to match what can be seen in a single pixel from the International Space Station with the actual measurements taken on the ground.

“What we’re doing is sampling trees inside those known pixels, measuring their gene expression and their spectral reflectance,” explained Swenson, a biological sciences professor and the director of UNDERC. “We’re doing that on the ground and trying to match that up with what the space station’s detecting—and testing how well we’re doing.”

The end game is to disentangle an individual NASA pixel from space into its parts on the ground of about 30 by 30 meters, which would make it possible to monitor drought signals from large swaths of forest. Compared to hand-measuring each tree in a land plot, the resulting models will mark a major leap forward in predicting forest health and potential wildfires.

Without braving the thirsty mosquitoes.

“The advantage here is instead of me out in the canopy with a little instrument measuring the chemical and structural content in a single leaf,” he said, “now you can fly a plane or satellite over the whole property and just like that, predict the chemical and structural content of every little pixel in the forest. And that’s been a pretty big revolution.”

Integrating emerging science

Swenson grew up camping in similar forests in Minnesota and was surprised to find in college that he “could get a job looking at trees all day.”

He was a forest ecologist at the University of Maryland when the pandemic started in 2020. He had been studying gene expression in the leaves of Eastern forests for years, noting what leaf genes do when they are happy with plenty of snowmelt and rain during the May-June growing season. This changes in July and August when there is more heat and less rain, and the color and reflectance of the leaves respond to drought.

“Our lab is one of the few in world to look at how genes are expressed in a natural forest for many species, rather than just one species,” he said. “We developed methods and analysis to compare gene expression across species and time.”

Under pandemic confinement, he read scientific papers about new developments in remote sensors taking spectral images from planes or space. Spectral imagery collects information beyond the visible spectrum by analyzing a range of wavelengths to reveal chemical and structural makeup along with the spatial information in an image.

He realized there was an opportunity to integrate this emerging field with his genetic research on the ground. He proposed a grant to NASA that would match their spectral imagery with his genomic work to predict forest health. It was a revolutionary but untested idea.

Around the same time, a group of Japanese researchers published a study about spectral signatures in rice fields. Drone flights measured spectral signatures to see how a single rice plant gene was responding to phosphorus in the rice paddies, proving that spectral imagery could be used this way on a landscape scale.

“The big quantum leap we'll make now is—instead of one species in a monoculture with a single gene—we’ll be looking at 20 or so species in a mixed stand of trees,” Swenson said. “And somewhere around 20,000 to 40,000 genes that are expressing themselves in a leaf at any given time.”

He acknowledged that the project means big data analytics with both plant genomes and spectral imaging. NASA was interested but only decided to fund the project after a leading DNA sequencing platform company, Illumina, stepped in to help support the research. The project will run for three years with just over $1 million in funding.

In 2019, Swenson saw the job opening at UNDERC in the type of forest he grew up in and knew that Notre Dame's facilities and students offered world-class opportunities. The 7,500-acre property with 30 lakes is one of a kind. He arrived at the place he would direct a day before teaching his first class there.

“I was amazed, when you drive through the main gate and keep driving for over a mile, and then you see the labs—it's incredible that the University has this kind of facility,” he said.

He also knew UNDERC was part of several larger studies with overlapping data. Through a partnership since 2008 with the National Ecological Observatory Network, a plane flies over the property taking spectral images in three of every five years. UNDERC is also part of a Smithsonian Institution ForestGEO program of long-term forest monitoring around the world, where nearly 80 research groups use standardized methods of tagging, measuring, and identifying tree species.

In June, Swenson and Rubio gathered samples from nine trees in each of 15 random plots at UNDERC, as well as another 15 plots at the Chequamegon–Nicolet National Forest an hour south. The process will be repeated in August and again next year to increase scientific accuracy.

“We try to do about five plots a day, so it takes about three days to complete,” said Rubio, who is from Colombia and completed her doctoral degree at Notre Dame in May. “Working on this project is helpful for my future because I’m learning new techniques for ecology.”

Land O’Lakes gem

Swenson said the team, facilities, and previous work at UNDERC make the research possible. But it actually started with controlled experiments in the Galvin Life Science Center on campus. In a greenhouse, Swenson grew trees in two groups, with a control group and one that is deprived of water to create stress.

“What we measure in the greenhouse and predict from images from space, is it any good?” Swenson said. “You don’t have an answer until you go out and measure it. So we’re doing that now.”

Ongoing ForestGEO work has tagged about 40,000 trees at UNDERC, essentially any tree that's over an inch in diameter. Hunter Gonzalez, a junior from southern Illinois, was part of that effort last year and this summer is doing an independent research project.

He and another student wearing special jackets with mosquito net hoods were spending the day examining trees for heart rot in their cores. In a circle around the trunk, they pounded in sensor spikes connected by wires to a sonic tomograph. By hitting each spike in succession with a mallet, the tomograph creates a picture like an MRI of the trunk interior.

“I'll probably get 20 trees done in a good, productive day,” said Gonzalez, a first-generation student whose passion for rivers morphed into tree research. “I'm used to the mosquitoes.”

While the work isn't specifically connected to Swenson’s research, it will provide overlapping data by showing whether the trees in the spectral images suffer from disease.

Swenson also teaches the forest ecology unit of a 10-week course at UNDERC for about 30 undergraduate students, half from Notre Dame and half from other schools. Other units include aquatic, vertebrate, and insect ecology.

In an hour-long morning class, Swenson taught the students how to use forest composition to learn its history. Different species have varying strategies for survival, with some trees racing to the canopy top for light and other longer-lasting types biding their time in the understory. Afterward, the students went out in the forest to identify maple or basswood from cedar or ash, and then check their answers with a phone app called iNaturalist.

Michael Parent, a sophomore from the Boston area, explained how to find Ostrya virginiana, an ironwood tree that has one runt leaf in a group with two larger ones.

“We'll take this sample and measure its fluorescence on a device and put that in a graph to see how it's growing,” Parent said. “Fluorescence is based on growth rate, so we can see if it’s doing well enough to survive.”

While UNDERC in the summer is a hotbed of environmental science research, the property also has a prominent place in Notre Dame lore. Rev. Theodore Hesburgh, C.S.C., a former University president, loved to fish the lakes and sometimes convened meetings there when a thorny problem required focus and seclusion.

In July 1959, Father Hesburgh convened the evenly split members of the U.S. Commission on Civil Rights at UNDERC and managed to forge an agreement on 12 recommendations that became the foundation of the 1964 Civil Rights Act. President Dwight Eisenhower, shocked by their agreement, reportedly told Father Hesburgh, “We've got to put more fishermen on commissions and have more reports written at Land O'Lakes, Wisconsin.”

In 1967, Father Hesburgh convened Catholic educators there to discuss the role of the Catholic university in relation to the Catholic Church. The still-debated outcome promoting intellectual freedom and autonomy is often called the Land O’Lakes Statement (PDF).

Professor George Craig, entomologist and vector biologist, started bringing students to the property in 1976. Since then, the family of Bernard “Jerry” Hank ’51, a benefactor and family friend of Craig's, has made extensive improvements and donations, including a fund to pay the students doing research and study there.

Now, UNDERC research could help ecologists monitor forest health and show where conditions need mitigation to prevent wildfires.

“We should be able to have something that the Forest Service and other property managers could look at and say, ‘We know from molecular and reflectance information that these places are lighting up as particularly drought stressed,’” Swenson said. “What we don't know is how far we can scale this out once we have information on this forest. We just don't know yet.”