Female preferences for more elaborate signals are an emergent outcome of male chorusing interactions in túngara frogs

In chorusing species, conspecific acoustic interference exerts strong selection on signal form and timing to maximize conspicuousness and attractiveness within the signaling milieu. We investigated how túngara frog calling strategies were influenced by varied social environments and male phenotypes, and how calling interactions influenced female preferences. When chorusing, túngara frog calls consist of a whine typically followed by 1–3 chucks. In experimental choruses we saw that, as chorus-size increased, calls increasingly had their chucks overlapped by the high-amplitude beginning part of other callers' whines. Playback experiments revealed such overlap reduced the attractiveness of calls to females, but that appending additional chucks mitigated this effect. Thus, more elaborate calls were preferred when calls suffered overlap, though they were not preferred when overlap was absent. In response to increasing overlap risk in larger choruses, males increased call elaboration. However, males overwhelmingly produced 2-chuck calls in even the largest choruses, despite our results suggesting additional chucks would more effectively safeguard calls. Furthermore, aspects of male phenotypes predicted to limit call elaboration had negligible or uncertain effects, suggesting other constraints are operating. These results highlight how complex interrelations among signal form, signaling interactions, and the social environment shape the evolution of communication within choruses.- R markdown - Jupyter notebookFunding provided by: University of Texas at AustinCrossref Funder Registry ID: http://dx.doi.org/10.13039/100008562Award Number: Integrative Biology department travel/research awardFunding provided by: National Science FoundationCrossref Funder Registry ID: http://dx.doi.org/10.13039/100000001Award Number: IOS-1914646Data have been collected by recording individual male frogs in choruses of different sizes. Timestamps of these calls were generated with custom python scripts. Most of the analysis uses data on the intervals between calls, the number of notes per call, and the temporal relationships between calls. Additionally, results from female binary choice tests (did she choose stimuli 1 or stimuli 2) are also included