The chirps and trills haunting the Jurassic forests of Netflix’s new Steven Spielberg–produced series The Dinosaurs weren’t simply dreamed up by Hollywood sound designers. Some were reconstructed from fossil insect wings 165 million years old.
Now, the research behind that soundtrack has been published in the Proceedings of the National Academy of Sciences. By combining exquisitely preserved fossil wings with experiments on living insects, computer simulations, and machine learning, scientists have re-created what they say is a partial Middle Jurassic soundscape. If the reconstruction is correct, one of the insect calls would be the oldest known evidence of ultrasonic communication in animals—long before echolocating bats appeared.
“It was really fun to read,” says Robin Tinghitella, a behavioral ecologist at the University of Denver who studies the evolution of cricket communication but was not involved in the work. “Just the idea that you could actually listen to this hypothetical soundscape that they generated in the paper was really cool.”
Sound itself, of course, doesn’t fossilize. But in crickets and katydids—also known as bush crickets—some of the anatomy that produces it can. A serrated vein, or “file,” on one of the forewings of these insects is lined with tiny, toothlike projections. As a scraper on the opposite wing runs across them, parts of the wing vibrate and produce sound. The length and spacing of those projections, along with the size of the vibrating wing area, provide clues to the pitch and structure of the resulting call.
In the new study, researchers analyzed 20 fossils representing nine species that lived at roughly the same place and time in what is now Inner Mongolia, allowing them to build an ensemble of animals that shared the same Jurassic environment. The scientists first tested their biomechanical models on living crickets, katydids, and their close relatives, using lasers to measure how their wings vibrate. They then applied the validated models to the fossil wings and checked the results against call frequencies predicted from the living insects. Machine learning helped estimate how rapidly the extinct insects repeated their basic sound units.
Most of the Jurassic species probably trilled at relatively low frequencies, comparable to many modern crickets. Ultrasound is commonplace among modern katydids, but its evolutionary origins have been much harder to pin down. One katydid relative, Sigmaboilus peregrinus, appears to have produced calls between 20 and 22 kilohertz—just above the upper limit of human hearing.
Many of the reconstructed calls were also unusually narrow in frequency—in other words, a pure tone. “A pure tone is a musical tone,” says Fernando Montealegre-Zapata, a sensory biologist at the University of Lincoln and one of the study’s lead researchers. To human ears, the audible calls would have sounded like clear, high-pitched whistles or electronic beeps, repeated in cricketlike chirps, rather than the raspy buzzes or hisses produced by some insects today. Such calls concentrate most of their energy in a tight frequency band.
That may have helped the insects communicate in a crowded nocturnal soundscape while making them harder for predators to pinpoint. “You hear a cricket singing, but try to localize it,” Montealegre-Zapata says. “You really can’t.”
That possibility gives the findings an evolutionary twist. Bats, often invoked as a major force driving insects toward ultrasonic communication, did not appear until roughly 100 million years later. Instead, the researchers suggest, early mammals or mammallike predators may already have been eavesdropping on insect courtship calls. It’s “an absolutely plausible idea,” Tinghitella says. “There are lots of things that are listening in on insects.”
Montealegre-Zapata and his colleagues picture an ancient acoustic arms race: As early mammals and their relatives became better at locating singing insects, selection may have favored purer or higher pitched calls, while the insects’ own hearing evolved to keep pace. Competition among species for uncluttered frequency channels may have added another pressure, effectively dividing the Jurassic night into acoustic niches—some close to the ground, others higher in the vegetation.
The reconstruction has limits. Fossil anatomy can reveal the pitch and basic sound an insect produced, but not the rhythm and the temporal pattern of its song, even though the latter is “really important for communication in singing insects,” Tinghitella says. The pattern is controlled by the nervous system and wing movement, neither of which is preserved in these fossils.
Still, these extinct insects are unusually revealing. Most dinosaurs produced sounds using soft-tissue organs that rarely fossilize, making their calls much harder to reconstruct.
Woven through multiple episodes of The Dinosaurs, the reconstructed soundscape changed the way Montealegre-Zapata imagines the ancient world. Jurassic forests, he says, were “a very busy acoustic environment”—a chorus extending into frequencies no human visitor could have heard.