Many migrating animals, including insects, rely on a celestial compass to navigate. They mostly use UV and polarised light, which are invisible to humans. While waves of unpolarised light are orientated in many different planes, some sunlight is scattered by the atmosphere and becomes polarised,meaning that the waves oscillate in one plane only. Scientists use a term called the E-vector to describe the plane of orientation of a polarised light wave, perpendicular to the direction the light wave is travelling. A group of neurons in the locust's brain, called POL neurons,respond to polarised light, but Uwe Homberg from the University of Marburg,Germany, wanted to know how the neurons would respond to different light wavelengths, too. To start categorising POL neurons' responses, Homberg and his colleagues Michiyo Kinoshita and Keram Pfeiffer focussed on two large and easily accessible POL neurons called LoTu1 and TuTu1 in locusts' brains(p. 1350).

To find out which light stimuli the two neurons responded too, the team recorded their electrical activity using microelectrodes while shining polarised and unpolarised lights of different wavelengths, and from different directions, onto one of the locusts' eyes. They found that both neurons had different and varied responses to the different types of light: green, blue and UV, either polarised or unpolarised.

The team found that LoTu1 responded to unpolarised green light shining from the side onto the eye, but shining unpolarised UV light from the same direction stopped all activity. When the team shone unpolarised blue light onto the top of the eye, there was no activity in LoTu1, however when the light was polarised, the neuron responded but the strength of the response depended on the E-vector orientation. This means that the neuron would respond more strongly when the sun was in a certain position relative to the locust.

The team found that TuTu1 responded in the opposite way to LoTu1 to light shining from the side: it responded to unpolarised UV light, but unpolarised green light inhibited the response in most experiments. However TuTu1 responded in a similar way to LoTu1 to polarised blue light, shone onto the top of the eye, firing most strongly at specific E-vector orientations. When they shone unpolarised blue light on the locusts' other eye, activity in the neuron stopped, suggesting that signals from the opposite eye can block neuronal signals in TuTu1.

The team were surprised that both neurons responded to unpolarised and polarised light. `We wouldn't expect a neuron to be sensitive to different colours and intensities' says Homberg, `it would interfere with the polarisation signal'. However, relying on polarised light alone means that an insect can't tell if the sun is to its left, or to its right. Being able to respond to unpolarised light as well as differences in the sky's colour would pinpoint the sun's position and solve this dilemma. Another possible advantage is that the responses to polarised and unpolarised light would be used at different times of day. When light levels are low, and the insect can't see the sun, the polarised response could dominate. However if the sun was visible, the unpolarised response would be more reliable. The insects are probably `combining features of the sky' to tell them where they need to go,Homberg explains.

Kinoshita, M., Pfeiffer, K. and Homberg, U.(
). Spectral properties of identified polarized-light sensitive interneurons in the brain of the desert locust Schistocerca gregaria.
J. Exp. Biol.