Western biologists have developed an innovative way to reconstruct how crickets sing, based on the physical formation of the chirping insects' wings, using measurements from preserved samples and computational modelling.
The new best practices, published July 30 in Royal Society Open Science, were devised by Western biology professor Natasha Mhatre, Canada Research Chair in invertebrate neurobiology, and three former undergraduate students in her lab, which investigates the biophysics of insect and spider communication.
Natasha Mhatre
In the new study, Mhatre and her collaborators detail a new computer modelling method that adheres more closely to a cricket's actual physical characteristics than previous attempts. The new model can predict the precise vibration patterns of cricket wings, even those of new wings that the model was not based on.
Scientists like Mhatre often use preserved specimens for a deeper understanding of evolutionary history and genetics. But recreating how extinct or dead birds and mammals, including humans, once sounded is complex. Both communicate using a vocal tract, intricately controlled by the brain. But both structures are made of soft tissue, which rarely fossilizes or leaves a trace. Crickets, however, sing a different tune - literally and figuratively.
Cricket songs are not a vocalization at all but are created by vibrational mechanics within their forewings. These tough, leathery wings, located in front of the hindwings, act as protective shields for crickets. They also house specialized, hardened microstructures needed for chirping. (Their hardness means forewings preserve well as fossils or museum specimens.) Most importantly, the venation pattern, or arrangement of veins, within the cricket forewing determines its song frequency or pitch.
"Each cricket wing has a pattern of veins running through it, which are structurally critical to making songs," said Mhatre, a biology professor in Western's Faculty of Science. "Some of these veins are used to generate the forces that make the wing vibrate and make sounds. Others stiffen local areas within the wing and develop the resonant structures that vibrate at specific frequencies."
What's the frequency, cricket?
For years, Mhatre and others in the global neurobiology research community have attempted to use bioacoustics (sound produced by living organisms) and finite element modeling (a method for numerically solving differential equations) to understand cricket song with a primary goal of predicting wing vibration and sound production. Thousands of cricket wings are preserved in museums and their evolutionary relationships are clearly mapped out, making this tactic for predicting their sounds a perfect pathway to unlock the mysteries of signal evolution and how some of the first sounds on earth sounded.
Teleogryllus oceanicus (Caroline Harding)
In fact, Mhatre and her collaborators thought they had cracked the code in 2012 in a game-changing study published in PNAS, in which they used some simple assumptions to develop computational models for cricket wings.
"There is a high density of veins in cricket forewings, so we considered these parts of the wing effectively immobile in our model. And this approach has stuck around for more than a decade," said Mhatre. "But something about this approach has always bugged me."
The issue with earlier studies, including the 2012 PNAS paper, is that the modelled cricket wings were 'clamped' at points with a high density of veins and not just at the base, as wings are hinged in nature. This simplified the study but technically, these parts of the wing are free to move, so the computer model wasn't a direct representation.
"We also don't really have an objective means of deciding what a 'high density' really means, in terms of veins in a cricket wing, which is a problem if you start with a new cricket with different wing venation whose wings you have never measured before," said Mhatre.
In the new study, Mhatre and her collaborators developed a computer modelling method that adhered more closely to the cricket's actual physical characteristics and clamped the wings as they should be.
The new model, based on Teleogryllus oceanicus (commonly known as the Australian, Pacific or oceanic field cricket). is now able to predict the precise vibration patterns of cricket forewings without simplifying assumptions. It can even predict the behaviour of new wings that it was not specially designed or tuned for.
The authors then tackled another method used for reconstructing cricket song, which was the use of preserved specimens. They showed that a dry-preserved cricket forewing, such as a museum specimen, would have very similar vibrational pattern as a live cricket but it would resonate at the wrong frequency. This is because the wing material hardens as it dries. They found that the correct frequency could be recovered, however, simply by wetting the wing with water or reducing this stiffness artificially in a computer model.
"We've developed a more reliable way of dealing with reconstructing cricket acoustic function from morphology, using computational modelling and preserved specimens," said Mhatre.
Mhatre collaborated with co-author Nathan Bailey from the University of St. Andrews on this new model, research that dates back to the beginning of the COVID-19 pandemic. The long-time collaborators were joined by three of Mhatre's then-undergraduate thesis students, Sarah Duke, Ryan Weiner and Gabriella Simonelli, who are all co-authors on the study.
