How Physics Can Replenish Birds’ Ancestral Songbooks

Biologists and neuroscientists want to know what songs mean and how they’ve evolved. For example, songs are usually about territorial defense or courtship. It’s interesting to know why certain themes have become established and why some are sung more than others. There are many hypotheses. One is a numbers game. Many individuals sing it, so, the theme survives over time. But there’s also the question of whether there’s something in a species’ neural architecture that makes certain combinations or motor gestures preferable. The traditional approach to studying questions like these is by examining brain activity during learning.

As a physicist, I turn that around and study the problem of birdsong not from the brain down, but from the bottom up. I asked myself a question many years ago: Could the acoustic and timbral properties of a bird’s song arise largely from the interactions between a set of genetically encoded instructions and a sophisticated physical mechanism? This led me to formulate the first vocal production models, which account for the movement of the membranes that modulate airflow in a bird’s respiratory system.

For years, I had studied how different animal species generate distinctive timbres through biomechanical properties and the interaction of such properties with the nervous system. For example, different bird species produce similar low-frequency sounds by engaging different sets of muscles. My colleagues and I measured the activity of those tiny muscles and the pressure of the air sacs. It was quite challenging because we had to take measurements when the animal is exhibiting behaviors that indicate different “thought patterns”—a good mood trying to find a mate, or a defensive mood trying to protect its territory. That requires implanting electrodes surgically—which takes a lot of skill.

We can learn about songs that a bird’s recent ancestors might have sung, even though these songs no longer exist in that bird’s repertoire. For example, in an experiment that arose during the pandemic when we couldn’t go to the lab, we wanted to “restore” a species’ song. That is, could we teach it songs that had been sung by its predecessors decades ago or in different parts of the world? The experiment took place in Pereyra Iraola Park, 60 km away from Buenos Aires. We were researching the species present in the park and found that the Andean sparrow (Zonotrichia capensis) was the most ubiquitous.

We recorded and catalogued the sparrow’s different songs and compared them to spectrograms from a paper published in the 1960s—a shorthand record of the calls and their meanings. There were hits from the ’60s that were still being sung and others that weren’t. Our idea was to place acoustic robots in an area frequented by juveniles who hung out together, like teenagers in a park, to popularize one forgotten song.

Yes! The robots were programmed to broadcast the song from three different locations at certain times, as if three adults were singing. Toward the end of summer, we used these programmable electronics also to record responses. We found that a juvenile had picked up the broadcasted song. The next spring, when that individual was an adult, recordings indicated that it had learned the robots’ song and made it its own.

It had a distinctive feature: The second syllable has a decreasing frequency modulation, which is very easy for us to identify and was not there before.

I work extensively with ornithologists from Argentina and around the world. I contribute my knowledge of how a bird generates the properties of its song. We see thousands of species that have unique and unusual strategies, and we encounter nontrivial physical phenomena all the time. For example, my biologist colleagues and I came across a certain pigeon that had an unexpectedly raspy timbre. We predicted it had asymmetry in the syrinx—an avian vocal organ that splits air at the base of the windpipe. Postmortem inspections of the bird’s vocal apparatus revealed asymmetrical lips—an anatomical validation of what we proposed in the model.

When you think about biodiversity and conservation, you think about genetics. But birdsong is learned and taught, and therefore we can talk about culture. Urban development has reduced the number of birds in Pereyra Iraola Park over the past 50 years. Along with that dwindling population, there’s been a cultural impoverishment. Behavior that once existed is gone. Our work helps raise awareness that biodiversity is more than genes. It’s culture: knowledge that has been transmitted and has been successful for survival. That’s what I love about it!