Acoustic differentiation of calls in the closely related Dartford (Curruca undata), Marmora’s (Curruca sarda) and Balearic Warblers (Curruca balearica)

ABSTRACT

Acoustic differentiation in bird vocalisations is a key process for speciation. Calls have received little attention compared to songs, but their simplicity facilitates analysis using mathematical, non-biased, tools, as they can be described by means of very basic variables. Here, we combined PERMANOVA and machine learning techniques to analyse the differentiation in the contact calls of the closely related Dartford (Curruca undata), Marmora’s (C. sarda) and Balearic (C. balearica) warblers (formerly in the genus Sylvia). The calls of the three species differed clearly, consistent with phylogenetic reconstructions based on DNA sequences. In C. sarda, significant call differences were observed between the population wintering in Tunisia (migratory) and those of Sardinia and Corsica (mainly sedentary). Such divergence could be favoured by assortative mating of migrant and resident individuals, working as a matching mechanism for reproductive isolation in sympatry. In contrast, there was no clear separation between the accepted subspecies of C. undata. We found, however, significant differences between the Atlantic and Mediterranean biogeographical regions in this species. The divergences detected within C. undata and C. sarda deserve further investigation, using additional traits to elucidate the phylogenetic status of these groups, as well as the factors driving their diversification.

KEYWORDS:

• Acoustic behaviour
• contact call
• geographic variation
• insularity
• machine learning
• Xeno-canto

Acknowledgements

Thanks to all the contributors to the open-access repository Xeno-canto. We also thank two anonymous reviewers for their constructive comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All the bird sound recordings used in this study are available at https://xeno-canto.org/. The codes for each recording are specified in Table 1.

Additional information

Funding

This work was supported by the Ministerio de Ciencia, Innovación y Universidades, Spain [PGC2018-096656-B-I00].