Machine learning reveals relationships between song, climate, and migration in coastalZonotrichia leucophrys

Abstract

ABSTRACTVocalization behavior in birds, especially songs, strongly affects reproduction, but it is also highly impacted by geographic distance, climate, and time. For this reason, phenotypic differences in vocalizations between different bird populations are often interpreted as evidence of lineage divergence. Previous work has demonstrated that there is extensive variation in the songs of White-crowned Sparrow (Zonotrichia leucophrys) throughout the species range, including between neighboring (and genetically distinct) subspeciesZ. l. nuttalliandZ. l. pugetensis. However, it is unknown whether the divergence in their songs correlates to environmental or geographical factors. Previous work has been hindered by time-consuming traditional methods to study bird songs that rely on the manual annotation of song spectrograms into individual syllables. Here we explore the performance of automated machine learning methods of song annotation, which can process large datasets more efficiently, paying attention to the question of subspecies differences. We utilize a recently published artificial neural network to automatically annotate hundreds of White-crowned Sparrow vocalizations across two subspecies. By analyzing differences in syllable usage and composition, we find thatZ. l. nuttalliandZ. l. pugetensishave significantly different songs. Our results are consistent with the interpretation that these differences are caused by the changes in syllables in the White-crowned Sparrow repertoire. However, the large sample size enabled by the AI approach allows us to demonstrate that divergence in song is correlated with environmental difference and migratory status, but not with geographical distance. Our findings support the hypothesis that the evolution of vocalization behavior is affected by environment, in addition to population structure.LAY SUMMARYBirdsong is an important behavior because it is important in bird communication and reproduction.White-crowned Sparrows in western North America are known to use different songs along their range, but it is unknown if those songs vary due to the environment.We used machine learning to analyze these songs and found that populations of White-crowned Sparrows can be differentiated based on their songs.Environmental factors during the breeding season exert a greater influence on song evolution in migratory subspecies.