PROCESSING OF AMPLITUDE-MODULATED SIGNALS THAT MIMIC ECHOES FROM FLUTTERING TARGETS IS THE INFERIOR COLLICULUS OF THE LITTLE BROWN BAT, MYOTIS-LUCIFUGUS

1. Neurophysiological recordings were undertaken to determine how neur ons in the central nucleus of the inferior colliculus (ICc) of the lit tle brown bat, Myotis lucifugus, extract amplitude modulations that sp an across a series of tone pulses (i.e., signals that simulate echoes from fluttering targets). Two types of stimuli were presented to the b ats. The first served as a control and consisted of an unmodulated tra in of tone pulses having different repetition rates (TPu, 5-400 pulses per second). The second was a train of tone pulses that were sinusoid ally amplitude modulated (TPm, 5-110 Hz) across sequential pulses. The modulated trains of pulses were presented at five different repetitio n rates (25, 50, 100, 200, and 400 pulses per second) encompassing the range of biosonar emission rates in these bats at different stages of target-directed flight. 2. One hundred fifty-two single neurons were isolated in the ICc of M. lucifugus; their basic response properties a nd temporal firing patterns were characterized. The best frequencies ( BFs) ranged from 10 to 80 kHz and the minimum thresholds at BF were di stributed widely (10-95 dB SPL). The frequency tuning selectivity rang ed widely, from very broadly tuned (Q(10db) = 1.3) to narrowly tuned ( Q(10dB) 89). Units with very narrow frequency tuning (Q values > 20) w ere restricted to BFs of 30-50 kHz. The temporal firing pattern of ICc units could be categorized into primary-like (PL), chopper (C), onset -immediate (O-I), and onset-late (O-L). 3. In response to TPu ICc unit s exhibited varying degrees of response selectivities as evidenced by their count-based response functions (using the spike count as a measu re) versus repetition rate. The count-based response functions of ICc units exhibited five filtering characteristics including band-pass, lo w-pass, high-pass, band-suppression, and all-pass characteristics. The temporal firing pattern of a unit showed certain correlations with it s count-based response function. For example, the majority of O-I and O-L units, and about half of the C units, showed tuned band-pass respo nse functions. The remaining C and onset types showed mostly low-pass response functions. In contrast, PL neurons showed mostly high-pass re sponse functions, but one third displayed band-pass response functions . 4. The ability of ICc neurons to time-lock their discharges to the i ndividual pulses in a train was characterized by using the synchroniza tion coefficient (SC) as a measure. The SC was plotted against the rep etition rate to construct units' synchronization-based response functi on. The vast majority of iCc neurons with FL, C, and O-L response patt erns showed low-pass synchronization-based response functions with the remaining manifesting allpass response functions. In contrast, the ma jority of O-I neurons displayed all-pass synchronization-based respons e functions showing effective time-locking across the entire range of repetition rates tested, with the remaining units showing low-pass res ponse functions. The average TPu cutoff frequency (the frequency at wh ich the SC dropped to 50% of maximum value) for low-pass neurons was, in order of decreasing time-locking ability: O-I (231 pulses per secon d), O-L (200 pulses per second), PL(110 pulses per second), and C (94 pulses per second). 5. ICc neurons were particularly sensitive to ampl itude modulations that were imposed across a sequence of sound pulses (TPm stimuli). Over half of the neurons, at all repetition rates teste d, showed a preference for a specific range of across-pulse AM frequen cy that corresponded to the range of insect wingbeat frequencies that little