Back

Lark Sparrows have Ecogeographic Song Variation across North America

Fuertes, S. H.; Provost, K. L.

2026-08-28 evolutionary biology
10.64898/2026.08.25.745997 bioRxiv
Show abstract

Machine learning models can be used to analyze large bioacoustics datasets and explore variation due to geography or habitat. We find that in the monotypic Lark Sparrow (Chondestes grammacus), both environmental variables and geographic distance influence song variation in this species. Bird song is an important method of communication within avian species. The variation in bird song within a species can be due to a variety of factors, including genetic drift and isolation by distance. However, it remains unclear in species with wide ranges how environmental factors in particular can cause changes to the song. In this study, the song C. grammacus was analyzed via machine learning to determine if it had significant variation based on multiple geographical metrics. We trained a convolutional neural network to segment individual syllables of 91 C. grammacus recordings, then extracted song characteristics. We found that ecoregion and state explain variation in C. grammacus songs. Our results demonstrate the efficacy of using machine learning models to analyze large datasets, as well as the impact that ecogeographic variation has on song variance.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.