Engineers bolted ultrasonic speakers to turbines at a Pennsylvania wind farm and searched the ground beneath them every day for two summers, and the broadcasting turbines killed up to half as many bats, because a bat avoids air it cannot listen through

The experiment ran on the least glamorous method in field biology: walking in circles, looking at the ground.

Every morning through the summers and falls of 2009 and 2010, search crews at the Locust Ridge wind farm in the ridge country of eastern Pennsylvania worked the gravel pads and mowed plots beneath a set of turbines, scanning for small furred bodies in the grass. Some of the turbines above them were ordinary. Others had been fitted with something new: boxes of ultrasonic speakers, aimed into the airspace around the blades, screaming at frequencies no human ear can register.

The question was simple. Would bats stay away from a machine they could not hear past?

Why anyone would yell at a bat

The idea sounds like a gimmick until you remember how a bat perceives the world. An insect-hunting bat flies by echolocation, firing off ultrasonic calls and reading the returning echoes to build a picture of the night, down to the wingbeats of a moth. Its ears are its eyes.

That is the vulnerability the deterrent targets. Broadcast enough ultrasound into a patch of sky, across the same frequencies the bats use, and the echoes drown. The bat is not frightened by the noise so much as blinded by it, flying into a region where its sonar returns nothing but glare. Certain moths evolved the same trick millions of years earlier, jamming attacking bats with clicks of their own. And crucially, bats do not like it. In earlier trials at ponds and in flight cages, bats abandoned foraging areas the moment the speakers switched on, and in feeding tests a bat never once managed to grab a tethered mealworm while the device was broadcasting.

The team behind the wind farm test, led by Ed Arnett of Bat Conservation International with collaborators from the US Geological Survey and Humboldt State, wanted to know whether the effect held around a 400-foot turbine with blades in motion, at a facility where the stakes were real.

Two summers of counting

The setup was clean: turbines randomly assigned as controls or as deterrent carriers, daily searches beneath both, and fatality estimates corrected for the classic field biases, the carcasses scavengers remove and the ones searchers miss. Wind farms in the eastern forests kill bats in numbers, particularly migratory hoary and silver-haired bats moving through in late summer, so there was no shortage of grim data.

The results, published in PLOS ONE, showed the speakers working, imperfectly but unmistakably. In 2009, treatment turbines killed an estimated 21 to 51 percent fewer bats than the silent controls. In 2010, once the team accounted for an inherent difference between the turbine groups, the range widened, from no real effect at one end to 64 percent fewer deaths at the other. Species mattered enormously: hoary bat deaths ran twice as high at control turbines as at broadcasting ones, with silver-haired bats close behind, while high-frequency callers like eastern red bats showed little benefit.

The pattern made physical sense. Ultrasound attenuates fast in air, the higher the frequency the faster it dies, so the speakers’ output thinned before it could fill the whole rotor sweep. Bats calling at lower frequencies, whose hearing overlapped the deterrent’s strongest output, got the message; bats operating above it flew on through. It did not help that the prototype devices kept failing in the field, taking on water and overheating through the study.

From prototype to product

For a first full-scale attempt with leaky hardware, halving fatalities for the worst-hit species was enough to keep the idea alive, and the years since have been a steady engineering grind. Later trials in Texas found deterrents cutting hoary bat deaths by 78 percent, and Brazilian free-tailed bat deaths by 54, while other deployments at Midwestern farms produced weaker or mixed results, with the same species-by-frequency split running through all of it. Manufacturers now build deterrents into nacelles at the design stage, and operators increasingly pair them with the other proven tool, curtailment, which idles blades on the low-wind nights bats prefer to fly.

That pairing is the practical answer taking shape across the industry: slow the blades when bats are busiest, and fill the air with jamming sound for the ones that come anyway. Neither is perfect alone. Together they attack the problem from both ends, the machine’s behavior and the animal’s senses.

The sound of an empty sky

There is something almost courteous about the mechanism, as wildlife protection goes. No fence, no light, no chemical. Just a message broadcast in the only language the animal fully trusts, saying, in effect, nothing can be known here. A bat that cannot sound out a volume of air treats it the way a person treats a pitch-black staircase, and turns away.

The bodies in the grass at Locust Ridge were counted so that fewer would fall elsewhere, and the tally they produced established a principle now bolted to turbines from Texas to the Great Lakes: the best way to keep a bat out of the blades is not to block its path, but to blind the sense that would have led it in.