On a scorching morning in early July, Julie Allen and her all-woman genetic sampling team of early career scientists and undergraduate students were ferried up the Rio Negro in Brazil. A pod of dolphins served as escort.

Allen’s group was a part of Limelight Rainforest, a multinational team from 10 institutions that traveled deep into the Amazon Rainforest to compete in the final round of XPRIZE Rainforest. Each of the six teams in the finals had 24 hours to autonomously survey the biodiversity of a patch of rainforest.

Automated technologies offer a gentler, more hands-off approach to biodiversity surveying that can help keep rainforests intact while significantly increasing the number of species surveyed, said Allen, a Virginia Tech biologist in the College of Science.

“Rainforest species are disappearing faster than we can survey them,” Allen said. “If we don’t have a fast way to survey forests, there won’t be forests left to survey.”

At the stroke of noon on July 8

In the shadow of a giant clock counting down the hours, Allen’s team successfully released a squad of drones equipped with traps to collect images, acoustics, insects, and other data streams across the rainforest canopy to survey biodiversity.

A major leg of the team’s strategy relied on environmental DNA, also known as eDNA, trace bits of DNA that organisms shed in soil, water, and even air. By simply brushing a leaf or sampling a scoop of water, scientists can assemble a snapshot of all the organisms in a given place.

As soon as it was deployed, the drone team run by Outreach Robotics began scouting for an on-site body of water, where eDNA can be found in high densities. The plan was to lower a water sampler through the canopy. But carefully. Dropping a sampler or losing a drone would violate the no-contact rule of the competition.

“It’s all about risk. How much do we trust this hole in the canopy to get our sampler down there and back up again?” Allen said. “Fortunately, we had remote sensing models that identified the most likely places to find water to reduce our search area.”

Growing urgency in the afternoon

But team members didn’t see water in the early afternoon. Or the late afternoon. Allen recalled her growing sense of urgency as the light lengthened.

Please, please, please let us find water, and be able to collect water,” Allen said. “But there was none.”

“Like none at all,” said Isabella Burgos, Virginia Tech lead DNA operations manager.

A sampler in a pool of water as seen through dense foliage
The team carefully lowered a water sampler through the rainforest canopy to collect environmental DNA. Photo courtesy of Outreach Robotics.

Evening win

Finally, as the sun sank low into the sky, an angled sunbeam sparked up the tiniest reflection through the canopy. What was it? A pool? A spring?

Allen never knew for sure, but it was big enough for a few samples, and team members got a good long pull. By the end of the day, they had located several sources of water and filtered enough for eDNA extraction.

Meanwhile, besides biting their nails off hoping for water, Burgos, Allen, and the rest of the sequencing team had been setting up a thick-walled, windowless eDNA tent to prevent sample contamination.

Brutal night

By the time the water and other samples got back, it was well after dark.

The eDNA team members, led by Virginia Tech postdoctoral fellow Niyomi House, worked in the enclosed tents through the night, wearing lab coats, masks, and gloves in the 95-degree heat, pipetting feverishly until dawn — all while being monitored by judges.

“It was brutal,” Allen said.

After it was all over, Burgos did the math: She slept 15 hours in six days, but it was “the best time.”

“Even if we sweated more than we had ever sweated in our lives and were beyond exhausted, we were still laughing, we were doing it together,” Burgos said.

Group of people smiling gathered around a table with their computers out.
The Virginia Tech-led genetic sampling team of all-female early-career scientists and undergraduate students. Photo courtesy of Taylor Schmitz.

Final sprint

Data collection wrapped up before noon the next day, followed by a 48-hour sprint to process the samples and match them up against a huge database of organisms known to exist in Brazil. The team generated more than 27 million DNA sequencing reads in 48 hours and identified hundreds of species, including a jaguar and several other elusive creatures via eDNA.

But everything team members found stayed in the rainforest.

“We brought nothing back with us. The sequences are being registered with Brazil’s National System of Genetic Resource Management and Associated Traditional Knowledge system,” Allen said. “This way, they know what we found, and the data stays in the country where it was produced.”

Because, Allen said, the competition is not so much about what they found as how they found it: a rapid big-picture survey of the rainforest that leaves no trace.

Limelight Rainforest is committed to abiding by Brazilian laws related to genetic heritage, associated traditional knowledge, and benefit sharing. XPRIZE collaborated with the National Institute of Amazonian Research to obtain necessary permits for collecting eDNA/DNA sampling activities during the testing days. All eDNA/DNA sequencing data will be registered through SISGEN, ensuring transparency and adherence to regulations, and no genetic data will be published without clearance from XPRIZE and its government partners, ensuring compliance with all relevant regulations.

Share this story