SOUND LOCALIZATION IN CRICKETS .2. MODELING THE ROLE OF A SIMPLE NEURAL-NETWORK IN THE PROTHORACIC GANGLION

Intracellular recordings were made from the auditory interneurons ON1, AN1 and AN2 in the prothoracic ganglion of the cricket Gryllus bimacu latus. Their responses to synthesized calling song (carrier frequency 5 kHz, intensity 40-90 dB SPL), presented monaurally and binaurally vi a legphones (acoustic trachea cut), were recorded (Fig. 2). These data were then analysed to determine the strength of inhibitory coupling o f ON1-ON1, ON1-AN1 and ON1-AN2 (Fig. 3). Inhibitory coupling of ON1-ON 1 and ON1-AN1 are relatively independent of stimulus intensity, ON1-ON 1 = -0.25 to -0,3 ap ON1(a)/ap ON1(b); ON1-AN1 = -0.4, to -0.55 ap AN1 /ap ON1(contra). The inhibition of AN2 by ON1 is rather ineffective, m aximally -0.13ap AN2/ap ON1(contra). The left-right contrast enhanceme nt or gain, due to lateral inhibition via omega neurons, is 1.6-1.9 fo r the ON1 pair and 2-3.4 for the AN1 pair, over most of the relevant s ound intensity range. For the AN2 pair, there is little contrast enhan cement except at sound intensities > 80 dB SPL (Table 1). The above da ta on auditory information processing in the prothoracic ganglion was incorporated into a simple model of the neural events underlying phono tactic behavior. This model included typical peripheral auditory direc tionality and simple assumptions about how phonotactic turns are gener ated. Predictions generated by the model include: a) the enhancement i n directional sensitivity of AN1 and AN2 provided by lateral inhibitio n (Fig. 4). b) The effects on open- and closed-loop phonotactic behavi or, of inactivating AN1 or AN2 (Fig 5). c) The effects of inactivating one ON1 on the directional sensitivity of left and right AN1 (Fig. 6) , and the consequent effects on phonotactic behavior (Fig. 7) Several of these predictions agree well with currently available experimental data, others constitute hypotheses for testing in future experiments.