Fig. 4.

Frequency-specific ITD tuning in the nOv and ICC of two device-reared owls. a, Data from the nOv of owl 13 (n = 20 sites) and owl 14 (n= 12 sites). Lines connect data collected at single sites. Negative ITD values represent left-ear leading; positive ITD values represent right-ear leading. In both owls, best ITDs changed with stimulus frequency in the adaptive direction (linear regression;p < 0.0001). b, Data from the lateral shell of the ICC in the same owls. Narrowband best ITD plotted as a function of stimulus frequency for sites tuned broadly for frequency. Lines connect data collected at single sites. In neither owl did best ITD change with stimulus frequency (p > 0.05). The value of ITD to which a site is tuned depends on its rostrocaudal location in the ICC.c, Broadband best ITD plotted as a function of best frequency for electrode penetrations through the lateral shell of the ICC (owl 13, 3 penetrations; owl 14, 8 penetrations).Lines connect data from the same penetration. The data on the left show a small dependence of best ITD on stimulus frequency (−3.4 μsec/kHz; p = 0.042;r² = 0.211), but the direction of this effect is opposite to the adaptive direction. The data on theright exhibit no dependence of best ITD on stimulus frequency (p = 0.966). The value of ITD to which a site is tuned depends on the rostrocaudal location of the recording site in the ICC.