An official website of the United States government
Oscine songbirds learn vocalizations that function in mate attraction and territory defense. Sexual selection pressures on these learned songs could accelerate speciation. The Eastern and Spotted towhees are sister species that diverged recently (0.28 Ma) but now have partially overlapping ranges with evidence of some hybridization; widespread community-science recordings of these species, including songs within their zone of overlap and from potential hybrids, enable us to investigate whether song differentiation might facilitate their reproductive isolation. Here, we quantify 16 song features to analyze geographic variation in Spotted and Eastern towhee songs and test for species-level differences. We then use random-forest models to measure how accurately their songs can be classified by species, both within and outside the zone of overlap. While no single song feature reliably distinguishes the two species, a random-forest model trained on 16 features accurately classified 89.5% of songs; interestingly, species classification was less accurate in the zone of overlap. Finally, our analysis of the limited publicly available genetic data from each species supports the hypothesis that they are reproductively isolated. Together, our results suggest that, in combination, small variations in song features may contribute to these sister species’ ability to recognize their species-specific songs.
more »
« less
This content will become publicly available on June 17, 2026
The role of learned song in the evolution and speciation of Eastern and Spotted towhees
Oscine songbirds learn vocalizations that function in mate attraction and territory defense; sexual selection pressures on these learned songs could thus accelerate speciation. The Eastern and Spotted towhees are recently diverged sister species that now have partially overlapping ranges with evidence of some hybridization. Widespread community-science recordings of these species, including songs within their zone of overlap and from potential hybrids, enable us to investigate whether song differentiation might facilitate their reproductive isolation. Here, we quantify 16 song features to analyze geographic variation in Spotted and Eastern towhee songs and assess species-level differences. We then use several machine learning models to measure how accurately their songs can be classified by species. While no single song feature reliably distinguishes the two species, machine learning models classified songs with relatively high accuracy (random forest: 89.5%, deep learning: 90%, gradient boosting machine: 88%, convolutional neural network: 88%); interestingly, species classification was less accurate in their zone of overlap. Finally, our analysis of the limited publicly available genetic data from each species supports the hypothesis that the species are reproductively isolated. Together, our results suggest that small variations in multiple features may contribute to these sister species’ ability to recognize their species-specific songs.
more »
« less
- Award ID(s):
- 2327982
- PAR ID:
- 10634978
- Publisher / Repository:
- Public Library of Sciecne
- Date Published:
- Journal Name:
- PLOS Computational Biology
- Volume:
- 21
- Issue:
- 6
- ISSN:
- 1553-7358
- Page Range / eLocation ID:
- e1013135
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
The dark-eyed junco (Junco hyemalis) has experienced rapid phenotypic diversification within the last 18,000 years, resulting in several subspecies that reside in partially overlapping regions across North America. These subspecies have distinct plumage and morphology. If members of a subspecies disproportionately mate with one another, we would expect genetic differences to accumulate between the subspecies. In parallel, their learned songs could also accumulate changes. If song is used by individuals to recognize members of their own subspecies during mate selection, which would prevent the production of less fit hybrid offspring between subspecies, then song differences might co-localize with subspecies boundaries. Here, we quantify 10 song features to explore subspecies-level song variation using song recordings from community-science databases. We build a machine learning classifier to measure how accurately the subspecies’ songs can be distinguished from one another. Here, we show that songs of dark-eyed junco subspecies exhibit significant song-feature differences. However, these differences do not necessarily lead to distinguishability between subspecies. Notably, we find that subspecies pairs with adjacent ranges that do not hybridize have much more distinguishable songs, and also more evidence for genetic differentiation, than pairs that are known to hybridize. Thus, song distinguishability appears to have predictive power about which subspecies will hybridize, suggesting that song might play a role in reinforcing certain subspecies boundaries more than others. Finally, we analyze subspecies-level song differences alongside available genetic data and geographic coordinates to characterize the current evolutionary landscape of the dark-eyed junco subspecies complex. We observe geographic signal in the song and genetic data, indicating that individuals who share a range are more likely to share song characteristics and be genetically similar. This study illuminates the existence of subspecies-level song differences in the dark-eyed junco and provides further clarity on the role learned song plays in reinforcing reproductive boundaries between dark-eyed junco subspecies.more » « less
-
Birds singing in choruses must contend with the possibility of interfering with each other's songs, but not all species will interfere with each other to the same extent due to signal partitioning. Some evidence suggests that singing birds will avoid temporal overlap only in cases where there is overlap in the frequencies their songs occupy, but the extent to which this behaviour varies according to level of frequency overlap is not yet well understood. We investigated the hypothesis that birds will increasingly avoid heterospecific temporal overlap as their frequency overlap increases by testing for a linear correlation between frequency overlap and temporal avoidance across a community of temperate eastern North American birds. We found that there was a significant correlation across the whole community and within 12 of 15 commonly occurring individual species, which supports our hypothesis and adds to the growing body of evidence that birds adjust the timing of their songs in response to frequency overlap.more » « less
-
Environmental stress, especially during development, can cause important phenotypic changes in individuals. In songbirds, these stress-induced changes have been shown to include impaired learning of song and reduced song complexity in several species. Typically, developmental stress has been studied in terms of individual variation within a population; however, birdsong, a culturally transmitted trait, may undergo relatively rapid changes if widespread learning impairment results from population-level environmental stress, and these changes could potentially be amplified if affected individuals become the song tutors for future generations. We thus hypothesize that ecosystem-wide stressors may cause population-level changes to birdsong. Here, we use publicly available birdsong recordings to determine whether song differences were evident after an abnormal and severe 2016 drought in western New York State. We analyzed birdsong recordings of two species, the Dark-eyed Junco (Junco hyemalis) and the Song Sparrow (Melospiza melodia), recorded between 2006–2020 in the drought-affected region and, for comparison, in two nearby regions less affected by the drought. The population-level song features of the species with more complex songs (Song Sparrow) changed in the drought area after 2016, but not in the control area. In the species with a more simple song (Dark-eyed Junco), we detected song changes in both regions, suggesting that the drought did not have an outsized effect on song in this species. These findings support a more nuanced hypothesis that stress-induced deficits may disproportionately affect species with songs that are more difficult to learn. These conclusions are tempered by the relatively sparse recording availability from years prior to 2016, but we predict that future longitudinal studies of song evolution in natural populations will be more tractable given the nearly exponential increase in the number of song recordings deposited in public repositories in recent years, making this experimental design a useful framework for future studies.more » « less
-
Learned traits are thought to be subject to different evolutionary dynamics than other phenotypes, but their evolutionary tempo and mode has received little attention. Learned bird song has been thought to be subject to rapid and constant evolution. However, we know little about the evolutionary modes of learned song divergence over long timescales. Here, we provide evidence that aspects of the territorial songs of Eastern Afromontane sky island sunbirds Cinnyris evolve in a punctuated fashion, with periods of stasis of the order of hundreds of thousands of years or more, broken up by evolutionary pulses. Stasis in learned songs is inconsistent with learned traits being subject to constant or frequent change, as would be expected if selection does not constrain song phenotypes over evolutionary timescales. Learned song may instead follow a process resembling peak shifts on adaptive landscapes. While much research has focused on the potential for rapid evolution in bird song, our results suggest that selection can tightly constrain the evolution of learned songs over long timescales. More broadly, these results demonstrate that some aspects of highly variable, plastic traits can exhibit punctuated evolution, with stasis over long time periods.more » « less
