When a male Anna’s hummingbird pulls out of a courtship dive at roughly 60 miles per hour, the air ripping through his outer tail feathers produces a loud, sharp chirp. It is not a vocalization. No syrinx is involved. The sound comes from the feathers themselves, vibrating like reeds in a wind instrument, and it is aimed squarely at the female perched below.
For more than a decade, UC Riverside biologist Christopher Clark has been recording, dissecting, and cataloging these mechanical sounds across hummingbird species. His lab’s work has established that hummingbird courtship is not just a visual spectacle but an acoustic one, with wing trills, tail chirps, and feather buzzes functioning as precisely tuned signals. Now, as Clark’s group expands its recording program and the broader field begins probing how hummingbird brains process sound, researchers are closing in on a question that has gone unanswered: whether the neural circuits birds use to interpret learned songs also decode the non-vocal sounds their bodies produce in flight.
Feathers as instruments
The foundational discovery came from a deceptively simple experiment. Clark and his collaborators filmed Anna’s hummingbird dives with high-speed cameras, then selectively removed individual tail feathers and sent the birds back up. When the outer tail feathers were gone, so was the dive chirp. Wind-tunnel tests of isolated feathers confirmed the finding: specific feathers vibrate at specific frequencies when exposed to airflow at dive speeds, producing tonal sounds as reliably as a tuning fork.
The physical mechanism behind this was later pinned down as aeroelastic flutter, a self-sustaining oscillation that occurs when airflow interacts with a feather’s shape and stiffness. This ruled out an earlier hypothesis that vortex shedding drove the vibrations. The distinction matters because aeroelastic flutter produces sounds at frequencies determined by feather geometry and material properties, meaning that even small evolutionary changes in feather structure can shift the pitch of a courtship signal. In effect, natural selection can tune the instrument.
Work on Calliope hummingbirds revealed that a single courtship display can layer simultaneous wing, tail, and vocal components, each operating at different frequencies. Males were observed shifting their wingbeat frequency during specific display phases, suggesting the wing trill is not a passive byproduct of flapping but something the bird actively modulates. Research on male Allen’s hummingbirds reinforced this, showing that wing trill production changes with flight speed through deliberate kinematic adjustments. The birds appear to amplify or suppress certain wingbeat tones depending on what they are doing, which points to a degree of voluntary control over their own mechanical soundscape.
A brain built for learned sound
Hummingbirds are one of only three bird lineages known to learn their vocalizations, alongside songbirds and parrots. That ability depends on specialized forebrain nuclei, and researchers have mapped several of these regions in hummingbirds using ZENK gene expression, a molecular marker that flags neurons activated during specific behaviors. The resulting maps revealed song-related brain nuclei strikingly similar to those found in songbirds, evidence of convergent evolution in neural architecture for vocal communication.
On the sensory side, behavioral and ZENK-based experiments have shown that at least one hummingbird species can detect high-frequency signals from its own kind, with corresponding activation in auditory brain regions. This confirmed that hummingbird hearing is tuned to the upper frequency ranges relevant for both vocal calls and the mechanical sounds their feathers produce.
What no one has done yet is connect these two bodies of evidence in a single experiment. The ZENK mapping studies tested responses to vocal signals. The aeroacoustic studies measured physical sound production. The question of whether the brain regions that decode learned songs also respond to wing trills and tail chirps remains open.
What Clark’s lab is building toward
Clark’s research program at UCR, described in a May 2026 university report, continues to expand the catalog of mechanically produced courtship sounds. The lab has used infrared high-speed video to capture scissor-tailed nightjars clapping their radius bones together during predawn displays, applying the same toolkit of synchronized audio, high-speed imaging, and controlled airflow analysis that underpins the hummingbird work. The broader agenda, as outlined on the lab’s research page, spans both sound generation and sound suppression, investigating how some species produce loud mechanical signals while others, like owls, have evolved feathers that fly in near silence.
The next logical step, and the one the field is converging on, would pair controlled playback of mechanical courtship sounds with neural activity measurements in listening birds. If wing trills activate the same forebrain song nuclei that respond to vocal calls, it would suggest hummingbirds process all courtship-relevant sounds through a shared neural pathway, regardless of whether those sounds originate in the syrinx or in a vibrating tail feather. If different circuits light up, it would mean the brain distinguishes between vocal and mechanical signals in ways researchers have not yet imagined.
Why the gap between production and perception matters
For now, the evidence reshapes how biologists think about hummingbird courtship even without the neural link fully established. Tail spreads and wingbeats are not silent visual flourishes. They are precisely tuned acoustic signals, generated through physical mechanisms that evolution can act on feather by feather. The sounds are repeatable, species-specific, and under at least partial voluntary control, all hallmarks of a communication system rather than incidental noise.
But communication requires a receiver, and the receiver’s brain is where meaning gets assigned. Until experiments trace a continuous pathway from feather vibration through the auditory system to a perceptual or behavioral decision, the story remains half-told. The spectacular aerial performances have been decoded mechanically. The sensory and neural logic that makes them meaningful to the bird watching from a branch below is the piece still missing.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
