Some problems are so universal that even insects' brains must be able to solve them. Male crickets croon a species-specific song, and female crickets meet up with attractive-sounding males by following a song to its source. A cricket's ears (located on the front legs) connect to just two thoracic interneurons and two more neurons ascending into the brain. Thus, an invertebrate's ‘neural parsimony’ leads to single neurons performing tasks that require whole neural subsystems in mammals. Merely within the thorax, the biophysics of the ears combine with local comparisons to quickly identify the song's frequency and its direction. How steering maneuvers follow from that information still remains something of a mystery, so Maja Zorovic and Berthold Hedwig took up the challenge to find out what kind of processing occurs in the brain to transform these auditory signals into motor commands.
These experiments were technically challenging. In order to show that a given brain cell was involved in phonotaxis, Zorovic and Hedwig had to show that the neuron both responded to a calling song and altered its activity during walking. However, recording from and identifying a neuron must be done by impaling its axon with a hollow glass electrode, sharpened until the tip is approximately 10 μm across. Such a tenuous connection is highly sensitive to vibrations or movements – like those occurring in a walking animal! The authors ameliorated this motion by affixing the cricket to a metal pin and suspending it above an air-floated plastic ball, so that the cricket could move the ball with its legs but its body would remain stationary. They further stabilized the brain by immobilizing the head, cutting away some of the surrounding muscles and sandwiching the brain into a tiny metal holder. The motion of the ball was then monitored using the sensor from an optical computer mouse.
In keeping with the theme of neural parsimony, Zorovic and Hedwig found just three brain cells processing the auditory information ascending from the thorax. These, in turn, fed a small pool of at least two neurons descending back to the body. The neurons along this pathway successively increased in selectivity to the correct song and in the robustness of their responses during walking, and nearly all the neurons showed a strong correlation with the walking movements of the cricket on the ball. Thus, it appears that a complete behavioral circuit can be mapped onto a chain of as few as five neurons, each uniquely identifiable in every individual cricket.
But why is the brain needed for this behavior in the first place? As sound frequency and direction are computed in the thorax, couldn't motor neurons be activated directly? The ability to integrate relevant information from other neural systems into the control of behaviors provides a clear adaptive advantage to an animal. In the case of crickets, obstacles, predators, hunger or thirst, or having already mated could all make a female cricket decide not to walk toward a calling male. Brains are where such high-level decisions take place – cutting out the brain would be taking the concept of neural parsimony too far.