The Mechanosensory Hair Bundles of the Inner Ear Distinguish Sinusoidal Forces from Noise Best When They Oscillate Spontaneously but Remain on the Verge of Quiescence

Hair bundles are the sensors that detect mechanical stimuli in the hearing and balance organs of vertebrates. Because these sensors are mechanically active, they often exhibit limit-cycle oscillations. A general theory of active hair-bundle dynamics predicts that a bundle's behavior may be controlled by its mechanical load. We confirm these predictions experimentally by employing a feedback system to change the mechanical load of individual hair bundles. In experiments, we observe that a noisy bundle’s average displacement and entrainment in response to frequency-detuned sinusoidal forcing peak when the bundle spontaneously oscillates near supercritical or subcritical Hopf bifurcations. Owing to two distinct mechanisms intrinsic to the bifurcations, we also observe stochastic resonance in a bundle’s displacement and entrainment, which can be eliminated by changing the bundle’s operating point. Because the mechanism for stochastic resonance depends on the bifurcation type, noise allows us to distinguish the two kinds of Hopf bifurcation. The results detailed here apply to all detectors possessing a Hopf bifurcation.