Applying network analysis to birdsong research

Abstract

By studying animal vocalizations such as birdsong, our capacity to record and interpret acoustic data has opened many opportunities for objective studies of animal behaviour, song evolution, physiology and ecology. However, the analysis of such data sets is often complex, and can vary between research teams, study species and theoretical approach. We test the use of network analysis in categorical birdsong data sets, focusing on four main applications: (1) quantifying and describing patterns of song sharing between individuals and/or populations; (2) understanding factors driving, and the implications of, an individual's position within a song-sharing network; (3) analysing (song or syllable) transfer patterns; and (4) investigating how environmental factors influence song-sharing networks. This exercise provides an excellent example of the advantages of sharing methodologies across disciplines within behavioural ecology and will hopefully inform future studies focused on different aspects of song sharing and cultural evolution by providing new tools and techniques for analysis.