A test of the nonlinearity hypothesis: Is noise any different than deterministic chaos?

Abstract

Noisy, unpredictable sounds are often present in the vocalizations of fearful and stressed juvenile animals across many taxa. A variety of structural characteristics, called nonlinear acoustic phenomena, which include subharmonics, rapid frequency modulations, and deterministic chaos are responsible for the harsh quality of these vocalizations. Exposure to nonlinear sound can elicit increased arousal in birds and mammals. Past experiments have used white noise to test for effects of deterministic chaos on perceivers. However, deterministic chaos differs structurally from white noise, and unlike white noise (which is random energy at all frequencies) may differ dramatically depending on how it is produced. In addition, the subtle structural variation of chaos may not be distinguishable in the environment due to the attenuation and degredation of sound over distance and different habitat types. We designed two experiments to clarify whether American robins (Turdus migratorius) and warbling vireos (Vireo gilvus) discriminate between white noise and deterministic chaos. We broadcast and re-recorded white noise and two exemplars of deterministic chaos at 1, 10, 20, 30, 40, and 80 m across open and forested habitat and compared spectrogram correlation among stimuli. Spectrogram correlation results indicated that sounds degraded similarly in both habitats when compared to reference distances of 1 m. Comparing pairs of stimuli across distance suggested that Chua chaos was more easily distinguishable from noise and logistic chaos. In addition, all stimuli became more distinctive over increased distance. The second experiment tested behavioral responses of robins and warbling vireos to control sounds of tropical kingbird (Quiscalus mexicanus), white noise, and two exemplars of deterministic chaos (Logistic waveform and Chua). For both species, there was no overall effect of stimulus and 95% confidence intervals indicated some patterns in difference of proportion of time spent in relaxed behavior and looking. These results suggest that species responded to all stimuli but did not discriminate among them. White noise and deterministic chaos could be potentially perceived differently by birds but this is dependent on the structure of the deterministic chaos. Increasing sample size may provide stronger results since effect sizes are small. Also, using an alternate control stimulus may aid in resolving whether birds are able to discriminate among noise and chaos. Whitaker 3