Necks craned upward by the millions on Aug. 21, 2017, to see the moon cruise in front of the sun. Nine in 10 adults in the U.S. watched last year’s total solar eclipse, a viewership that beat every Super Bowl and the Apollo moon landing. But humans weren’t the only ones with their eyes on the skies. Wildlife reacted in rapid and sometimes striking ways.

We know this because ecologists, who have had about a year and change to analyze and publish their data, are sharing the results of studies they conducted during the eclipse. Birds cleared out of darkening air. Bees in the Midwest ceased buzzing, cut silent by a celestial conductor. And something, possibly flying insects, rose like a plume when the moon’s shadow was deepest.

Scientists of all stripes studied the eclipse. “You suddenly have a cue that is so unambiguous and so strong,” said postdoctoral researcher Cecilia Nilsson, a behavioral ecologist at Cornell University in New York who monitors bird migration via weather radar. Light, she said, is a tremendous cue to animals. The eclipse altered that cue on a massive scale.

There’s a long, if not especially deep, history of observing wildlife during eclipses. A 1935 paper rounded up reactions from the general public in New England, the site of an eclipse in 1932. Several people described unusually noisy crickets. Beekeepers reported bees returning to hives in droves. Toads swarmed a New Hampshire garden. A Massachusetts woman blamed the eclipse for a sudden cockroach infestation in her pantry.

Nearly a century later, ecologists can rely on more sophisticated tools than eyewitness accounts. Nilsson and her colleagues, as they report in a study published Tuesday in the journal Biology Letters, tracked birds and insects from 143 Doppler radar stations across the country. Eight of the stations were directly along the 2,500-mile path of totality.

Using machine-learning programs, ecologists like Nilsson can pick out the animal signals collected in weather radar. “We try to clear out as much weather as possible because it messes up our bird data,” Nilsson said. (Meteorologists do the opposite.) Radar offered them an overview of anything in flight from low to the ground to about three miles high. They were not able identify individual organisms in this study, nor could they distinguish birds from large flying insects, but their data revealed the total number of objects in the air.

For much of the day Aug. 21, the birds and bugs seemed as busy as usual. About 50 minutes before totality, though, animals began to vanish from high altitudes. “We were quite surprised to see this decrease of activity,” Nilsson said. Humans would struggle to notice the change in sunlight intensity so early.

She suspects those birds and insects sought shelter. Perhaps they reacted to the impending eclipse as they would to a storm. “They might be thinking that it’s heavy cloud cover coming in, which would also explain them landing, taking shelter, trying to figure out what’s going on,” she said.

At five of the sites, during the moment of totality, animals burst into action at low altitudes, only to settle down again. Though Nilsson was unsure how to interpret these bursts, she offered one hypothesis: Perhaps insects that feed at dusk rose into the air, thinking it was night, then became confused and landed. Birds might explain this plume, too; 70 percent of U.S. birds, including many songbirds, migrate under the cover of darkness. But Nilsson doesn’t think birds became so discombobulated as to disembark for a long flight.

University of Missouri ecologist Candace Galen, who was not involved in the radar report, applauded the authors for creating a “big-picture pattern of airborne biological signals.” This work, which included birds and bugs, captured “all the instruments in the orchestra,” Galen said. Galen’s own study focused on a single instrument: the bees.

Galen enlisted elementary school students and others to monitor bees during the eclipse. They placed tiny thumb-drive microphones — “we call them USBees,” she said — in flower patches in Missouri, Idaho and Oregon. In late summer in those states, garden are typically alive with the sound of bees. The sounds they recorded are not for communication, Galen said, but flight. The pollinators' wings are noisy, in a similar way to a jet engine’s roar.

The eclipse halted bees’ wings. Galen, in a study published last month in the Annals of the Entomological Society of America, found that bees abruptly stopped buzzing during the moment of totality. She had expected a gradual decline in bee activity. Instead it was a cliff. Just as soon as the sky brightened, they began to buzz again.

An eclipse is a once-in-a-lifetime event for the typical critter, Nilsson said: “Most animals will never have experienced this before.” From an astronomical perspective, total solar eclipses aren’t all that rare, visible from Earth every 18 months or so. From a evolutionary one, however, it’s hard to see how organisms could have adapted to handle eclipses. Paths of totality cross the planet at all sorts of directions and angles; world maps of eclipses look as jumbled as knife marks on a cutting board.

Scientists, though, can plan accordingly. Both of the researchers are anticipating the next total eclipse to pass through the U.S., which will cross from Texas to Maine in 2024. Nilsson wants to collect more precise radar data, which should allow her to determine the kinds of animals in the plumes. Galen will ready her tiny microphones. She hopes some of those same students will help her listen to the bees again, too.

The Great American Eclipse united the country for a brief moment of wonder and memes. But it appears to have confounded the heck out of the nation’s animals.

Read more:

Do eclipses drive animals wild? Here’s how you can help scientists find out.

2017 solar eclipse was one of most-watched events in American history, survey finds

‘We are all just mind-boggled:’ scenes from the total solar eclipse of 2017