Abstract
The auditory stimulation method used in experiments on moth A cell(s) is generally believed to be adequate to characterize the encoding of bat echolocation signals. The stimulation method hosts, though, several biases. Their compounded effects can explain a range of discrepancies between the reported electrophysiological recordings and significantly alter the current interpretation. To test the hypothesis that the bias may significantly alter our current understanding of the moth's auditory transducer characteristics, papers using the same auditory stimulation method and reporting on either spiking threshold or spiking activity of the moth's A cells were analysed. The consistency of the reported data was assessed. A range of corrections issued from best practices and theoretical background were applied to the data in an attempt to re-interpret the data. We found that it is not possible to apply
a posteriori
corrections to all data and bias. However the corrected data indicate that the A cell's spiking may (i) be independent of the repetition rate, (ii) be maximum when detecting long and low-intensity pulses and (iii) steadily reduce as the bat closes on the moth. These observations raise the possibility that a fixed action pattern drives the moths' erratic evasive manoeuvres until the final moment. In-depth investigations of the potential bias also suggest that the auditory transducer's response may be constant for a larger frequency range than thought so far, and provide clues to explain the negative taxis in response to the searching bats' calls detection.