An adult zebra finch teaches a young songbird a mating song in the lab of Dr. Todd Roberts at UT Southwestern's O’Donnell Brain Institute.
Newswise — Humans aren’t the only ones who can fall for a talented crooner. Songbirds do, too, though the features of an alluring avian serenade have been a long-standing mystery for about a third of the 4,000 songbird species that sing only one song.
With the aid of artificial intelligence, UT Southwestern researchers have discovered what they could not decipher with their own ears: Spacing out syllables in a more complex song is the key to a girl’s heart -- at least for a zebra finch.
New research published in Nature used an AI program to analyze half a million zebra finch syllables to identify qualities that signal sexual attractiveness. The program, the Deep Avian Network, helped uncover a crucial characteristic of a successful song: Females seemed to prefer longer paths between syllables, songs that were more difficult to learn. Researchers then tested their findings with female birds by generating synthetic songs with longer and shorter paths.
The females flew toward the songs with more complexity.
Dr. Todd Roberts, a neuroscientist from UT Southwestern who led the study, says having the resources and mental acuity to learn these more difficult songs might be more attractive to potential mates.
“For decades, scientists have tried to understand how a single song could signal reproductive fitness and allow females of the species to select among male suitors,” Dr. Roberts says. “Our research suggests that female songbirds are assessing male fitness in a much more complex way.”
MULTIMEDIA ASSETS/INTERVIEWS: Audio clips of the various bird songs are available upon request, in addition to video of the songbird lab. Contact for interview request: [email protected]
RSS