Date of Degree
David C. Lahti
acoustic analysis; bird songs; house finch
Bird song is a powerful model system in behavioral biology, especially for learning and cultural evolution. Understanding the origins and maintenance of vocal diversity in nature is fundamentally important to acoustic biology. Here, we propose a large-scale, integrative population analysis of nearly 2000 songs of the house finch (Haemorhous mexicanus) (190 individuals) collected in western Long Island of New York during an interval of 37 years (between 1975 and 2012) to explore cultural change of bird songs. Generally, I have three aims in this study. In the chapter 1, to facilitate acoustic analysis for field recordings with ambient noise, of large sample sizes and with continuous variation, I introduce an algorithm (FinchCatcher), which extract signals from the background noise, summarizes songs as a series of spectral shapes and performs dynamic classification that optionally combine the advantages of hierarchical clustering and partitioning. I further test the algorithm on a geographical comparison of 240 songs of the house finch from eight localities, using previously published observations as ground truth for estimating transitions of song (or song element) sharing during the spread of the house finch in North America. In Chapter 2, I quantify spatial variation in the songs of the house finch in western Long Island in 2012. Previous studies have suggested large morphological and behavioural differences between house finches in the east and west coast, which may have profound impact on song evolution of eastern house finch. The result shows great acoustic variation across the sampling range, rather than finding discrete areas within which song and syllable repertoires are highly similar between individuals. Nevertheless, spatial differentiation was neither simply clinal change with geographic distance, nor discrete dialects. To further explore the mechanisms underlying cultural change in this study region, in Chapter 3 I perform a temporal comparative analysis of house finch songs spanning nearly four decades. Substantial cultural change is observed in terms of local song and syllable sharing, song complexity and song type consistency. In addition, not a single song type persists over time and half of syllable types defined in the past data are not present in the recent songs. These results illustrate the potential interplay between multiple drivers of spatial and temporal variation. This body of research has provided a framework for understanding the spatiotemporal variation in house finch song in Western Long Island, incorporating partial migration, population growth, relaxed selection by females on male song, and developmental stress playing interacting roles.
Ju, Chenghui, "Cultural Evolution In Natural Populations: A Quantitative Bioacoustic Analysis" (2015). CUNY Academic Works.