Acoustic Signal Amplitude Patterns: A Computer Simulation Investigation of the Acoustic Adaptation Hypothesis

Abstract

The acoustic adaptation hypothesis (AAH) predicts that vocalizations intended for unambiguous long range communication should possess amplitude modulation (AM) characteristics such that the temporal patterning of amplitude is minimally degraded (due to atmospheric turbulence and reverberation) during transmission through native habitat. Specifically, signals should possess rapid AM (trills) in open habitats and low rate AM (whistles) in closed habitats. To determine which of these amplitude patterns incurs less degradation from its two main components, reverberation and irregular amplitude fluctuations (IAFs), we constructed two synthetic 'source' signals, a rapid AM 'trill' and a low rate AM 'whistle', from pure tone frequency sweeps. We applied the degradation components independently, thus avoiding their complex interactions typical of field recordings. Signals were degraded by various echo treatments (modeling closed habitat reverberation) or by various amplitude decrease treatments (modeling open habitat IAFs). Results revealed that the difference in performance between signal types lies not so much in their average transmission quality, as in the variability of that quality. In closed habitats, whistled signals transmit with more consistent quality than trilled signals over biologically realistic echo delays. In open habitats, trilled signals transmit with far lower variability of quality than do whistled signals. The inherent redundancy of trills transmits information more effectively than whistles in open habitats. Our results show strong support for the AAH predictions regarding what type of signal structure is best suited for open or closed habitats. However, this support is based on variability in performance of signal types in different habitats rather than average transmission quality.