Auditory detection of paired pulses by a dolphin in the presence of a pulse jam

From behavioral studies of a bottlenose dolphin (Tursiops truncatus), the audibility thresholds were measured for a single pair of equal-amplitude pulses, i.e., clicks, presented to the dolphin in combination with a pulse jam. The pulse jam consisted of pairs of identical pulses with a pulse spacing τ_j within the pairs and a pair repetition rate f _j. Series of pulses were interrupted by a pause R>1/f _j, within which the pulse jam was absent while a pair of test pulses was supplied to one of the two channels at random. Each series had a duration T, and the total stimulation cycle was J=T+R. The dependence of the test pair detection threshold on the pulse spacing τ_j was studied at different fixed values of the pulse spacing in the test pair: τ_t=50, 100, 200, and 500 µs. Preliminary measurements performed with τ_j=τ_t=100 µs were used to adjust the parameters of the pulse jam. The threshold shift at τ_j=τ_t=100 µs reached 35 dB above the audibility threshold of the test pair in the absence of the pulse jam. On both sides of the point τ_j=τ_t=100 µs the thresholds decreased with varying τ_j to approximately 20 dB above the detection threshold of the test pair in the absence of the jam. However, in the course of training, the threshold curves gradually shifted downwards approaching the detection level of the test pair in the absence of the jam and becoming progressively flatter (the selectivity with respect to the pulse jam vanished). A decrease in the pause duration R restored the dependence of the test pair detection threshold on τ_j. In this case, a statistically significant maximum was obtained at τ_j=τ_t for τ_j within the critical interval (for τ_t<500 µs). Beyond the critical interval (for τ_t>500 µs), even with the smallest pause duration (R=15 ms), no dependence of the test pair detection thresholds on τ_j could be observed.