To the human ear, the buzz of the honeybee can sound like one unchanging hum. Yet a group of researchers hopes that decoding tiny variations in the noise could help halt the catastrophic decline in the world's honeybee population.

The researchers, led by a team at Nottingham Trent University in England, believe the changing sounds from a hive indicate swings in the bees' state of health and that high-tech eavesdropping could provide beekeepers with early-warning signals. Supported by a $1.8-million grant from the European Union, the scientists aim to analyze the buzz from 20 hives kept at a village in rural southeastern France in a five-year experiment that started earlier this spring.

Team leader Martin Bencsik has previously used sensors known as accelerometers to capture a distinct change in bee sounds before the phenomenon known as swarming, which is when the queen quits the hive, taking many of the worker bees with her. The challenge this time is to identify variations in the buzz that can be linked to disease, including colony collapse disorder—a mysterious ailment that has weakened colonies around the world. The researchers' key tool: industrial sensors designed to pick up subtle changes in vibration patterns. Embedded in the wall of the hive, miniature accelerometers will measure the vibrations in the honeycomb caused by the bees' activity and the sounds they create. With no ears, bees are generally thought to rely on vibrations—received through their legs—to communicate with one another.

Researchers in Nottingham will then analyze the data from the hives, using computer software to find telltale correlations between the buzz—its pitch, for example, or the interval between pulses—and the bees' health. Bencsik foresees a time when any worrying changes logged by the accelerometers will trigger an automatic wireless alert to the beekeeper, who can then take swift action.

The trouble is, scientists are still struggling to identify—let alone treat—the main causes of the bees' decline, which some studies have linked to pesticides. “There are lots of suggested treatments, but there is no evidence that any of them work,” says Simon Potts, a professor of biodiversity and ecosystem services at the University of Reading in England. Beekeepers must hope that Bencsik's diagnostic tools will come of age at a time when more effective treatments are available.