In the April 13 issue of the journal Science, neuroscientists from the California Institute of Technology report that an owl locates prey in the dark by processing two auditory signal cues to "compute" the position of the prey. This computation takes place in the midbrain and involves about a thousand specialized neurons.
"An owl can catch stuff in the dark because its brain determines the location of sound sources by using differences in arrival time and intensity between its two ears," says Mark Konishi, who is Bing Professor of Behavioral Biology at Caltech and coauthor of the Science paper.
For example, if a mouse on the ground is slightly to the right of a flying owl, the owl first hears the sound the mouse makes in its right ear, and a fraction of a second later, in its left ear. This information is transmitted to the specialized neurons in the midbrain.
Simultaneously, the owl's ears also pick up slight differences in the intensity of the sound. This information is transmitted to the same neurons of the midbrain, where the two cues are multiplied to provide a precise two-dimensional location of the prey.
"What we did not know was how the neural signals for time and intensity differences were combined in single neurons in the map of auditory space in the midbrain," Konishi says. "These neurons respond to specific combination of time and intensity differences. The question our paper answers is how this combination sensitivity is established."
"The answer is that these neurons multiply the time and intensity signals," he says.
Thus, the neurons act like switches. The neurons do not respond to time or intensity alone, but to particular combinations of them.
The reason the neural signals are multiplied rather than added is that, in an addition, a big input from the time pathway alone might drive the neuron to the firing level. In a multiplication, however, this possibility is less likely because a multiplication reduces the effects of a big input on one side.
It's not clear how the owl perceives the location of the mouse in the third dimension, Konishi says, but it could be that the owl simply remembers how far it is to the ground or how much noise a mouse generally makes, and somehow adds this information into the computation.
The lead author of the Science paper is José Luis Peña, a senior research fellow in biology at Caltech.
Contact: Robert Tindol (626) 395-3631