Pigeons are thought to navigate by reading the Earth's magnetic fields, but they might not carry their compasses in their beaks, as previously believed.

For years, scientists have believed that the secret to the pigeon's internal GPS lay in magnetized iron-rich nerve cells in its beak. This was backed up by experiments showing that the nerves throughout the skin of a pigeon's upper beak respond to changes in magnetic intensity and several papers describing the discovery of iron-containing structures in pigeon beaks. Earlier studies have shown that when magnets or magnetic coils are attached to the heads of homing pigeons, their homing ability is disrupted and the birds become confused.

But a new study casts doubts on this hypothesis.

When an international team of scientists used imaging technology for a closer look at pigeon beaks and brains, they found iron particles, and not just inside nerve cells. Instead, they found that the iron popping up on their scans was actually located inside macrophages -- the white blood cells that eat up pathogens and cellular debris. They reported their results in a paper appearing in the journal Nature on Wednesday .

What's more, the iron that the researchers found in the macrophages wasn't even magnetic, so it couldn't pick up on magnetic fields, according to Mark Lythgoe, a co-author of the new study who hails from the Center for Advanced Biomedical Imaging at University College London.

The iron found in the pigeon's head likely comes from the macrophage-mediated breakdown of red blood cells, the authors say. But since macrophages cannot send electrical pulses like nerve cells, it's very unlikely that the iron in them is involved in magnetic sensing.

"Even if they were magnetized, they couldn't send information to the brain," Lythgoe said in a phone interview.

The results, the authors say, require the re-evaluation of a number of behavioral studies.

Other scientists did point out that the Nature paper doesn't undermine the central theory of magnetically mediated pigeon migration, but just reopens the question of how, physiologically, the birds can do it.

Kenneth Lohmann, a University of North Carolina at Chapel Hill biologist unaffiliated with the study, says the finding is interesting but he's not sure that it will really change how scientists view the big picture of animal navigation in the long run.

"There is still considerable evidence that the receptors for the magnetic sense are somewhere in a bird's head and probably in the general vicinity of the beak," Lohmann says .

Roswitha Wiltschko, a scientist from the University of Frankfurt, was even more critical. She says there's still a lot of evidence pointing to magnetism-sensing structures in the beak, even if the authors couldn't find them with their methods.

"Many birds behave differently when the skin on their upper beaks is numbed," Wiltschko wrote in an email.

In behavioral experiments, pigeons that were normally sent off-kilter by a magnetic anomaly were not confused if their upper beaks were numbed with Lidocaine. Anesthetizing the skin of the upper beak also alters migratory birds' responses to magnetic pulses, according to Wiltschko.

"Lythgoe and his colleagues' conclusion that all theoretical considerations ... and behavioral studies must now be re-evaluated just because they were unable to find the respective receptors is ridiculous," Wiltschko said.

But Michael Walker, a biologist at the University of Auckland in New Zealand, says the Nature study corrects errors that were made by other scientists.

"What we see now is the self-correction process of science taking place," Walker said in an email.

There are other ways that pigeons could sense magnetic fields, according to Henrik Mouritsen, who studies sensing in migratory birds at the University of Oldenburg in Germany.

"There's very good evidence that birds have a magnetic compass in their eyes," Mouritsen said in an email.

Mouritsen was senior author on another Nature paper in 2009 that examined a possible alternative to the iron-rich nerve cell hypothesis. According to this theory, birds sense the directions of magnetic fields through special photopigments in their eyes. The key to this system is thought to be a molecule called cryptochrome, Mouritsen says.

In the 2009 Nature paper, Mouritsen and his colleagues tested this light-dependent theory using 36 European robins. They altered the brains of some of the birds, making lesions in an area of the brain thought to be where the magnetic field signal is processed. These birds could not navigate by magnetic field, but were able to use the sun and stars to orient themselves, according to the study.

David Keays, a researcher at Austria's Institute of Molecular Pathology who led the latest study, said in a statement that the mystery of how migratory birds use magnetism to navigate remains an intriguing puzzle.

"We have no idea how big the puzzle is or what the picture looks like," Keays said, "but today we've been able to remove those pieces that just didn't fit."