The Case of the Disappearing Homing Pigeons

Between 1968 and 1987, about 900 homing pigeons released at the Jersey Hill fire tower in upstate New York got lost, never to be seen again. Why couldn't they find their way home?

Between 1968 and 1987, about 900 homing pigeons released at the Jersey Hill fire tower in upstate New York got lost, never to be seen again. Why couldn't they find their way home?

William T. Keeton (Cornell University)

Between 1968 and 1987, Cornell professor William T. Keeton and his colleagues released thousands of homing pigeons from different points in upstate New York and then tracked them to see if they could find their way home.

Homing pigeons are famously good navigators, and, for the most part, the pigeons set sail in the right direction. But there was one route that caused them trouble: A 74-mile stretch from the Jersey Hill fire tower back to their loft at Cornell. Only 10 percent of the pigeons trying to make that journey ever made it home. The rest -- about 900 pigeons -- disappeared completely.

Except, that is, on August 13, 1969. On that date, the pigeons released at Jersey Hill flew right back to Ithaca with no problems. Down on the ground, Keeton and his team took meticulous notes about the weather and whatever else they could notice. Nothing seemed different from any other day. They were mystified.

"When Bill Keeton was first setting up his pigeon loft at Cornell, he got birds from local pigeon racers," Jonathan T. Hagstrum, a geophysicist with the U.S. Geological Survey told me. "They told him: Don't go to Jersey Hill. You'll lose your birds."

There was something going on at Jersey Hill, but what? And why was August 13, 1969, so special?

* * *
In these diagrams, the center of each circle represents the release point, Jersey Hill. Each dot represents the direction a pigeon flew (north is at the circles' tops). The dashed lines represent the correct direction homeward. Diagram A shows the totality of 81 experiments conducted with pigeons from Cornell between 1968 and 1987; Diagrams B, C, D, and E show specific dates. (Journal of Experimental Biology)

Bird navigation is not well understood, and the hypothesized cues that birds use to make their "maps" (or how they point in the right direction homeward) are many: gradients in inclination and intensity of the geomagnetic field, visual landmarks, or even atmospheric odors. Keeton and his colleagues hoped that somewhere in that patch of upstate New York they'd be able to find something that was sending these pigeons astray. Whatever it was, it was likely a key component of pigeon GPS.

One possible clue: laboratory tests conducted in the '70s found that homing pigeons can detect what are known as "infrasounds" -- sounds at a frequency below the range of human perception. Humans, roughly speaking, can't hear sounds lower than 20 Hz, give or take a little for the sounds' intensity and the differing abilities of humans. (You can listen to a 20 Hz tone here.) Pigeons, it was found, could detect sounds at least as low as 0.05 Hz.

The world is downright awash in infrasounds. They can come from anything: explosions, earthquakes, storms. But the infrasounds that may guide birds come from one place: the ocean. The Atlantic Ocean, whose coastline is some 200 miles from Ithaca, New York, is banging against the floor of the deep sea, and that energy travels through the continent. "The East Coast is going up and down about four microns every six seconds," Hagstrum told me. The pulsing land acts like the cone of a stereo speaker, creating a pressure wave -- infrasound -- that birds would be able to hear. Additionally, another type of infrasound waves travel from the ocean through the atmosphere, not the land, and scatter when they hit mountains, buildings, and other vertical features of the terrain.

(Unlike the sounds humans hear, infrasounds can travel thousands of miles in the atmosphere. For example, a tone at 1000 Hz is 90 percent absorbed over a distance of seven kilometers at sea level. At 1 Hz, it takes 3,000 km to absorb 90 percent. For a tone at 0.01 Hz, "the distance exceeds the circumference of Earth," according to Hagstrum.)

An infrasound laboratory at the University of Hawaii captures examples of these infrasounds and then re-renders them into the range of human perception (20 Hz to 20 kHz) through methods such as pitch shifting and compression. Here, to give you a sense of this invisible-to-us sonic landscape, is one of their "infrasound" clips, this one of the surf on Moorea, an island in French Polynesia.

Could these mysterious infrasounds -- emanating from the oceans -- be the signposts that guide birds around the wide Earth? And, if so, why were the pigeons at Jersey Hill unable to hear them?

* * *

Following Keeton's death in 1980, the database containing the results of his pigeon releases was made available to all researchers upon request. Hagstrum took that data, and added some more: weather and topography. "The things that really influence sound are temperature, wind speed, and wind direction," Hagstrum said. He continued, "Those three things and then: the terrain."


Using a software program called HARPA (Hamiltonian Acoustic Ray-tracing Program for the Atmosphere), Hagstrum was able to model the infrasound landscape for Jersey Hill, taking into account the topography of the place and the particular atmospheric conditions at the times of the releases. (For the weather data, Hagstrum used a NOAA database of twice-daily wind measurements.) "So you can create this virtual model of the atmosphere and the terrain on the day the birds were released," he explained.

Hagstrum looked at the infrasound waves as they would travel from the Cornell loft to Jersey Hill and he found something very strange: Infrasound waves coming from Ithaca and heading westward skipped right over Jersey Hill, as he reports in his paperin the most recent issue of the Journal of Experimental Biology. Infrasounds from other locations reach Jersey Hill, but the sounds that carried the specific mark of the Ithaca loft -- the ones that announced "Home: This way!" to the pigeons -- they were absent. For the birds of the Cornell loft, Jersey Hill was in an acoustic "shadow zone."

Acoustic shadow zones -- also known as "zones of silence" -- are a normal phenomenon: Sound waves bend up through the troposphere (the lower 10 km of the atmosphere), reach the stratosphere where the temperature gradient changes, and are bent back down. The space in between where the rays go up and where they come back down is mostly silent. This is why in big explosions, such as the eruption of Mt. Saint Helens, people closer in often don't hear anything, while those further out do. Released in a shadow zone, the Cornell pigeons were lost.

But why, then, on August 13, 1969, were they able to navigate so well? When Hagstrum modeled the atmospheric conditions from that date, he found an anomaly: Something -- a wind shear or a temperature inversion at about a kilometer's height -- bent the sounds back down. "It bent it down early enough that it hit at Jersey Hill," Hagstrum said.

No one on the ground would have ever noticed anything odd weather-wise. It was a perfectly sunny day. Keeton and his colleagues took careful weather notes -- all for naught: The weather on the ground was more or less irrelevant. "This was going on a kilometer above their heads," Hagstrum elaborated. "They would have been sitting there completely oblivious." By Hagstrum's calculations, this sort of condition would have happened about 5 percent of the time -- infrequently enough that Keeton would have seen its effects just once.

A shadow zone (combined with these atmospheric fluctuations) solves the mystery of Jersey Hill -- and the bigger mystery of pigeon navigation -- quite neatly. Unlike the other possible mapping cues, such as the geomagnetic field, acoustic conditions within the atmosphere can change quite rapidly, explaining why pigeons can navigate from point A to point B just fine one day, and fail completely the next.

To a human (or, at least, to this human), it seems impossible that pigeons can discern the right direction out of the waves of sound that pulse around the world. But, as Hagstrum sees it, a pigeon's ability to find the right infrasound signature is not all that different from our ability to spot our own homes by sight. "How do you walk into your house and not somebody else's house?" he said. "I mean, houses, especially in some housing tracts, all look the same, basically." But somehow, just like pigeons, we manage -- navigating homeward by the signs we're equipped to perceive.