Song analyses
We analyzed 50 recordings belonging to the five subspecies of T. ruficapillus and covering the entire geographic distribution range of the species (see Table S2). Recordings were in ‘wav’ format and were obtained from the Macaulay Library of Natural Sounds (Cornell Lab of Ornithology) and Xeno-Canto.
We generated and analyzed the spectrograms using Raven Pro 1.5 (http://www.birds.cornell.edu/raven). The conditions for the analyses were a 512 fast Fourier transform length with a 50% overlap, a Hamming window type, and a grayscale color scheme. We analyzed one song per individual choosing the one with the best signal-to-noise ratio. Songs in this species consist of a repetition of similar notes. The first note is the longest, followed by successively shorter notes, until the last note, which is longer and similar to the first one. We measured five variables: mean duration of the note, mean duration of the interval between notes, fundamental frequency (calculated as the difference in frequency between the harmonics and measured in the center of each harmonic), bandwidth of the note (difference between the maximum and minimum frequencies of the note) and the number of notes in each song (Figure S1). Variables were measured on the first note, then additionally averaged between the second, fourth and sixth note to obtain a measurement of the intermediate notes, and finally again on the last note.
To assess vocal variation we performed a PCA and analyzed the data with a one-way analysis of variance (ANOVA) followed by Bonferroni’s contrasts to assess differences among subspecies in their PC scores (all the principal component scores met the assumptions of homoscedasticity and normality). All variables were standardized previous to the analysis and we considered that variables were significantly correlated with a PC only when the module of their factor loading value was greater than 0.7 (Tabachnick and Fidell 2001).
Finally, we also performed an ANOVA using the original song variables, followed by Bonferroni’s contrasts, because this can provide a more detailed analysis of specific differences among subspecies.