Vocal and genetic variation between a land-bridge island and mainland populations of the Black-crowned Antshrike (Thamnophilus atrinucha)

Abstract

Island-mainland systems serve as natural laboratories for studying communication signals. We explored potential divergence in the song of the black-crowned antshrike between Gorgona Island, a land-bridge island off the Pacific coast of Colombia, and several mainland populations. We found that the perching song of this species, the so called loudsong, was distinctly different on Gorgona Island. This differentiation encompassed an increase in frequency bandwidth, a decrease in vocal performance, and a reduction in song duration. All are consistent with documented patterns of divergence observed in avian acoustic signals on oceanic islands. The observed distinctions in loudsong acoustics, including variations in frequency and temporal features, led to the inference that the population on Gorgona Island produces the least complex loudsong dialect across the entire species range. This pattern of how an innate vocalization manifests within a land-bridge island, which was connected to the mainland not long ago, lends support to the Character Release Hypothesis. To ensure an unbiased comparison, we considered genetic clusters and analyzed them by comparing ten acoustic signals between the land-bridge island and several mainland populations. Combined with the gene genealogy, this finding supports the validity of two geographic forms: Thamnophilus atrinucha atrinucha and Thamnophilus atrinucha gorgonae (island). However, they also reveal the existence of additional unrecognized dialects of loudsong that transcend current taxonomic classifications. We provide evidence that challenges the long-standing presumption suggesting minimal intraspecific geographic variation in the vocalization among non-vocal learning species. Significance statement Significant evolutionary implications exist for an intriguing pattern of variation of the loudsong behavior in the Black-crowned Antshrike over a land-bridge island. Our gene tree does provide evidence of the enormous haplotypic diversity accumulated in the trans-Andean region long after the final uplift of the northern Andes. The acoustic divergence of the loudsong suggests substantial geographic variation within a non-vocal learning suboscine. Genetic and loudsong divergence on a land-bridge island add to document that evolution of acoustic traits may be relaxed in islands as compared to those observed on the species-rich mainland. The island-mainland system is one of the few definitive biogeographic contexts for understanding evolution of acoustic behavior.