SINGLE-UNIT SELECTIVITY TO AZIMUTHAL DIRECTION AND SOUND PRESSURE LEVEL OF NOISE BURSTS IN CAT HIGH-FREQUENCY PRIMARY AUDITORY-CORTEX

1. The azimuth and sound pressure level (SPL) selectivities of single-unit responses recorded in primary auditory cortex of barbiturate-anesthetized cats were studied by the use of broadband noise bursts delivered in the free field from a moveable loudspeaker. The experiments were carried out with cats located inside a quasianechoic sound-isolation chamber. We studied 71 units with relatively stable response properties. All units were located in the frequency representation between 5.8 and 31 kHz. The data obtained for each unit were displayed as an azimuth-level response area, a contour plot that displays the distribution of response magnitude as a joint function of SPL and azimuth at O° elevation. From these, azimuth and level functions were obtained to derive descriptors of azimuth and level selectivity. 2. Sensitivity to sound-source azimuth was assessed from the modulation of the average function (average of azimuth functions obtained to each SPL of noise that was presented) for each unit. The sample was arbitrarily divided into a high-directionality (HD) group (66%) whose average azimuth functions had modulation values of ≥75% and a low-directionality (LD) group (34%). The distinction between HD and LD groups was made so that we could analyze the characteristics of units likely to be involved in the representation of sound-source azimuth. 3. There is an overrepresentation of the contralateral sound field and the midline in the sample of HD units. The preferred sector for each unit was defined as the range of azimuths within the frontal sound field throughout which unit response was ≥75% of maximum. Each unit was classified as either midline preferring (17%, the midpoint of the preferred sector, i.e., best azimuth, was located within 5° of the midline), contralateral preferring (60%), or ipsilateral preferring (23%). The ratio of contralateral- to ipsilateral-preferring units was 2.5:1. A higher proportion of units had best azimuths located in the 10° sector centered on the midline than in any other 10° sector of the frontal sound field. 4. In one animal, recordings were obtained at seven closely spaced sites in layer IV from single- and multiunit responses, which were narrowly tuned to both azimuth and SPL. The units located along a 1-mm length of an isofrequency strip were tuned to similar frequencies and SPLs but had five distinctly different directional preferences distributed throughout the entire frontal sound field. 5. Azimuth and level selectivity of single units are related to each other. Units whose level functions are strongly nonmonotonic belong exclusively to the HD category. On the average, HD units are more strongly nonmonotonic and more narrowly tuned to SPL than LD units. Within the HD group, the units that are narrowly tuned to azimuth are usually narrowly tuned to SPL. A simple model involving bilateral nonmonotonic level selectivities is capable of generating neural responses that exhibit high-directional sensitivity and narrow SPL tuning. 6. Neurons with similar azimuth and level tuning appear clustered together in AI. At three recording sites, the response properties of two different units were studied, and, in each case, they had similar azimuth and SPL selectivities. Furthermore, responses of small clusters of units often appeared as selective as single units to azimuth and SPL. Both observations suggest that neighboring neurons exhibit similar azimuth and SPL selectivities. 7. Response strength is related to unit directionality and nonmonotonic strength. HD units had significantly lower response strengths than LD units, and strongly nonmonotonic units had significantly lower response strengths than monotonic and weakly nonmonotonic units. 8. The responses of a minority of HD units to a particular sound direction are relatively independent of SPL over a 40- to 60-dB range as reflected in response areas with vertically oriented iso-response contours. This feature is reminiscent of response characteristics of single units in the lateral superior olive. SPL-independent azimuth selectivity, when it occurred, was generally seen for low-response magnitudes. At high-response magnitudes, some units exhibited an interaction between azimuth and level tuning such that the maxima of azimuth functions obtained at different SPLs varied in location. In 13% of the HD units, the pattern of interaction was quite systematic. Azimuth functions obtained with increasingly higher SPLs had peaks that shifted laterally away from the midline.