In urban and human-modified landscapes, animals face novel selection pressures resulting from differences in habitat structure and increased anthropogenic noise. Urban noise pollution can negatively impact songbirds because low-frequency noise often masks portions of birds’ mating signals and reduces signal transmission. Previous research has demonstrated that the songs of birds in more urban habitats have structural differences that can enhance signal transmission when noise is present. The majority of these studies have focused on species that deliver short, stereotyped songs and have limited repertoires. Gray catbirds (Dumetella carolinensis, family: Mimidae) sing long bouts containing imitated, improvised, and invented song elements, and therefore may have an increased ability to vary songs in response to noise. We hypothesized that aspects of developed habitats including loud anthropogenic noise and changes to land cover would impact catbirds’ song structural parameters, including song minimum, peak, and maximum frequency, frequency bandwidth, and entropy. We recorded and processed songs from 42 male catbirds and analyzed over 18,000 song elements from sites along an urban gradient from western Virginia to the Washington, DC metropolitan region. We quantified the urban intensity at each site-centroid based on percent canopy cover, percent impervious surface, and noise level. Song features such as minimum, maximum, and peak frequency increased significantly as noise levels increased, demonstrating that catbirds in noisier areas sing higher frequency songs compared to individuals in quieter habitats. Land cover variables also significantly predicted certain song features such as maximum frequency (impervious surface) or entropy (canopy cover). These structural differences in catbird song can limit the negative effects of environmental noise-masking, even for their long song bouts, and suggest that vocal mimics respond to anthropogenic noise. Future studies could investigate repertoire size and composition along an urban gradient and if these structural differences lead to functional consequences for the songs of vocal mimics.
Read full abstract