Does the Structure of Tropical Forests Shape Bird Song?

The acoustic adaptation hypothesis predicts birds will produce signals that propagate best in their native environment. We tested the hypothesis that songs of white-breasted wood-wrens (WBWW) (Henicorhina leucosticta) differ with forest structure. We selected WBWW because they are sedentary on their territories year-round. We used song playback to prompt singing behavior. We sampled at distance intervals greater than a WBWW territory along trails at La Selva Biological Research Station, Costa Rica. Recordings were made with a parabolic microphone and analyzed with Raven Pro software. We classified songs into categories. We measured temporal and frequency parameters for each category, based on three randomly selected examples. We measured temporal and frequency parameters for each song category from each bird we recorded, based on three randomly selected examples. After this we classified recordings according to structure as primary forest, secondary forest, or forest adjacent river. Primary forests have high canopies and less understory growth. Secondary forests have denser understory growth. Ambient river noise affects the sound environment of adjacent forests. We suspected that denser understory would select for song centered at lower frequency, and ambient river noise would select for song centered at higher frequency. Our results to date suggest that WBWW song in secondary forests has slightly lower central frequency than primary forests, bordering on statistical significance. We also noted unique introductory notes that may convey individually recognizable markers of breeding territories.