Researchers at the University of Michigan have employed artificial intelligence to identify a previously underappreciated pattern in the evolutionary history of passerines, a bird group comprising most songbird species. The analysis indicates that these birds experienced periods of rapid evolutionary change interspersed with slower phases, with several major bursts of evolution corresponding to significant shifts in Earth’s climate conditions.
The study involved examining skeletal measurements from more than 2,000 passerine species, accumulating over 170,000 individual measurements from bird specimens. This large-scale data collection was made possible through Skelevision, an AI tool developed collaboratively by researchers at the University of Michigan and New York University. The system photographs bird skeletons against a consistent scale and can measure 12 bones across a specimen’s skeleton in approximately 45 seconds, enabling researchers to digitize museum collections far more efficiently than traditional methods. The team scanned and measured more than 15,000 museum specimens, primarily from the U-M Museum of Zoology collections.
To analyze the collected data, lead researcher Jake Berv created a new statistical method called bifrost, which examines complete skeletal structures rather than individual bones in isolation. This approach allowed the team to reconstruct approximately 45 million years of passerine body-shape evolution. The analysis revealed a particularly pronounced period of rapid body-shape evolution around 35 million years ago, which aligned with the Eocene-Oligocene transition marked by intense global cooling. Additionally, a cluster of evolutionary slowdowns occurred around 15 million years ago, coinciding with another major geological event.
The researchers further tested their findings by analyzing the geographic distribution of bird species in their dataset. They discovered that birds inhabiting more extreme latitudes, where seasonal temperature variations are more pronounced, tend to evolve faster in terms of body shape than species living nearer the equator. This pattern, consistent across both geological time and modern geographic regions, suggests that environmental variability may significantly influence rates of morphological evolution.
The findings underscore the growing value of museum collections in modern scientific research and carry implications for understanding how species might respond to contemporary climate changes. The research was supported by Schmidt Sciences, the David and Lucile Packard Foundation, the Michigan Institute for Data & AI in Society, the Natural Sciences and Engineering Research Council of Canada, and the National Science Foundation. Results were published in Nature Ecology & Evolution.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI