LEVEL DEPENDENCE OF COCHLEAR NUCLEUS ONSET UNIT RESPONSES AND FACILITATION BY 2ND TONES OR BROAD-BAND NOISE

1. The responses of onset units in the cochlear nucleus of the anesthe tized guinea pig have been measured to single tones, two-tone complexe s, and broadband noise (BBN; 20-kHz bandwidth). The onset units were s ubdivided into three groups, onset-I(On(l)), onset-L (On(L)), and onse t-C (On(C)), on the basis of a decision tree using their peristimulus lime histogram (PSTH) shape and discharge rate in response to suprathr eshold best-frequency (BF) tone bursts. 2. PSTHs were constructed from responses either to single tones at a unit's BF or to BBN as a functi on of level. When sufficient sustained activity could be elicited from the unit, arbitrarily defined as >100 spikes/s, a coefficient of vari ation (CV) was calculated; the majority were characterized by a CV tha t was similar to transient chopper units (0.35 < CV < 0.5). First spik e latency decreased monotonically with increasing sound level. For the majority of onset units, the first spike timing was very precise. 3. BF rate-level functions recorded from On(L) and On, units did not show any signs of discharge rate saturation at the highest sound levels we have used (100-115 dB SPL). No systematic relationship was observed b etween the threshold at BF and the shape of the rate-level function. B BN rate-level functions were typically characterized by higher dischar ge rates than in response to BF tones. However, for On(l) units and a minority of other onset units, there was little difference in the shap e of their rate-level functions in response to BF tones or BBN. 4. The threshold of most onset units to BBN was similar to the threshold to a BF tone that had similar overall root-mean-square (RMS) energy. The BBN threshold was on average, 5.5 dB greater than the BF threshold. Th is result contrasts with that found in auditory-nerve fibers recorded in the same species, with the use of an identical sound system, where the threshold to BBN was, on average, 19.4 dB higher. The mean thresho ld difference between BBN and BF tones for a population of chopper uni ts recorded in the same series of experiments was 17.7 dB. The relativ e thresholds to BBN and BF tones indicated that the bandwidths near th e onset units' BF threshold were broader than could be estimated with the use of single tones. Tell units were characterized by bimodal resp onse areas. 5. When the stimulus was two tones, one varying in level a t the units' BF and the other fixed in level at a frequency/intensity combination that was nonexcitatory by itself, the response was greater than the response to a single BF tone. For some units the threshold t o the two-tone complex was lower than the threshold to the BF tone alo ne. If the tone was within 0.5 octaves of BF, this threshold shift was on average 7.6 dB. Not all frequency/intensity combinations were capa ble of reducing the BF threshold. 6. These results show that the input to the onset units is from auditory-nerve fibers with a range of char acteristic frequencies even wider than chat implied by response areas defined with single tones. This is consistent with their enhanced temp oral response to the pitch of complex sounds and suggests that they ha ve sacrificed high resolution in the frequency domain for enhancement in the temporal domain.