Fig. 2: Anti-phase preference for call onset timing in broadband noise.

A Schematic representing the translation from the cartesian to polar coordinates. 0 phase (right) corresponds to the amplitude trough. Red trace represents amplitude change over a cycle. B Average call onset distributions within the modulation cycle track the inverse of the amplitude phase in broadband masker conditions, but not in silent and high-freq masker conditions (1 ms bins). Shaded areas indicate standard error of the mean (SEM). Black curves: schematic of amplitude envelope (not to scale) in playback conditions and simulated amplitude envelope in silent conditions. Insets: Circular density histograms of call onsets (30 bins). White arrows indicate resultant vectors: arrows indicate angular means; arrow lengths indicate resultant vector lengths (concentration of data at the angular mean). Shaded areas indicate maximum likelihood (MLE) 95% confidence intervals for angular means. Stars indicate significant p-values from Rayleigh’s test of uniformity. *** p < 0.001 C Bootstrapped MLE von Mises mean (μ) and concentration (κ) parameters indicating robust pattern of call onset clustering in the broadband masking condition for both modulation rates. D Data points (lightly colored dots) indicate number of calls observed per hour for each group for each recording day. Estimated mean calling rates (darkly colored dots with gray outline) from negative binomial models show a drop in calling between silent baseline and acoustic masking conditions (except for 15 Hz high-freq masker condition). Gray vertical lines indicate SE of model fit for predicted means.