Harvard University COVID-19 updates

Department News

What’s in a Squeak? Wild Mice Provide New Insights into Why Vocalization Evolves [Hoekstra Lab]

What’s in a Squeak? Wild Mice Provide New Insights into Why Vocalization Evolves [Hoekstra Lab]

All babies cry, but, as any parent can tell you, not all cries are the same. Across mammals, young animals are born with the ability to produce cry vocalizations that elicit care from their parents, and there are big differences in how much baby mammals cry and what they sound like when they do. In work now published in Current Biology, researchers in the Hoekstra Lab asked how such differences in vocalization have evolved. 

To do this, they focused on a unique group of rodents called deer mice. Deer mice are unique because large differences in traits (like vocal behavior) have evolved between species that are still so closely related they remain interfertile in the lab (even if they don’t interbreed in the wild). This makes deer mice an ideal group to understand both why and how behavior evolves. 

To investigate how vocal behaviors have evolved, the team—Nick Jourjine, Maya Woolfolk, Nacho Sanguinetti, Emory Sabatini (‘22), Sade McFadden, Anna Lindholm, and Hopi Hoekstra a.k.a, “Team Squeak”—started by recording the vocalizations produced by infant mice, called pups, when isolated from their nest. This is a useful context to study vocal behavior because these vocalizations are relatively easy to elicit and have been well studied in the standard laboratory mouse. However, relatively little was known about how other rodent species use these calls or how they have diversified within and between species. 

Led by postdoc Nick Jourjine, the team recorded isolation calls from almost 600 pups belonging to eight different deer mouse taxa (representing four different species) and compared them to the calls of both standard laboratory mice (Mus musculus) and a population of wild, free-living Mus musculus (living in a barn near Zürich, Switzerland). Specifically, they isolated pups into sound-attenuating boxes either in the lab (for deer mice and laboratory Mus) or the field (for wild Mus) and recorded the isolation calls each pup made in this context. They then used recently developed computational bioacoustics approaches to automatically detect vocalizations in the 128 hours of audio recordings they generated and cluster those vocalizations together based on similarity in their acoustic features, such as frequency and duration. 

Nick Jourjine, with undergraduate Emory Sabatini, found that unlike both wild and laboratory populations of Mus musculus, whose pups made isolation calls exclusively in the ultrasonic range (~65 kHz, well above the upper limit of ~20 kHz detectable by humans), all the deer mouse pups made two categorically distinct vocalization types: one in the ultrasonic range that resembled those made by Mus musculus, and one in the audible range (< 20 kHz). The latter has acoustic features that resemble cries made by many mammals, including human infants.  By comparing vocalizations across deer mouse species within each of these two vocalization types (ultrasonic vocalizations – or “USVs” – and cries), they found differences in the rate and acoustic features of both types, demonstrating that variation in pup isolation calls can evolve rapidly even between closely related species.  

The team also found similarities among species. They found that cries and ultrasonic vocalizations differ not just in how they sound (e.g. their frequency and duration), but in how pups use them. By looking across a pup’s first few weeks of life, they learned that shortly after birth–before they were able to thermoregulate, open their eyes or walk–pups primarily made cries, but once the pups were less reliant on their parents for care, the pups switched to primarily emit ultrasonic vocalizations. This pattern raised a hypothesis: perhaps the cries of young pups were better able to attract the attention of parents than USVs, when parental care was critical. To test this idea, postdoc Nacho Sanguinetti and grad student Maya Woolfolk played back the sounds of cries or USVs to deer mouse mothers and measured how the mothers responded to each vocalization type. They found that while both types of vocalization could elicit the attention of mothers, cries elicited much faster responses, consistent with the hypothesis that they are signals of urgent need by young pups. These experiments suggested that these two vocalization types may serve different functions. It’s tantalizing to imagine that while cries are more effective at eliciting rapid care, they may also be more detectable by predators, leading to the switch to USVs once the pups are older. 

While the team was starting to uncover why vocal types evolved, they also wanted to know how they evolved. For example, what are the respective genetic contributions (if any) to the differences they observed between deer mouse species? To answer this question, the team took advantage of two deer mouse species that are interfertile and thus can be bred to produce a hybrid population in which each individual hybrid has a genome that is a unique mixture of each species. By recording isolation calls from hundreds of hybrids, they found that the genetic contributions to interspecific variation in cries and USVs are largely controlled by separate regions of the genome (i.e. different genes). This result suggests that the evolution of deer mouse pup cries and USVs are not constrained by one another at the genetic level, meaning that they can evolve independently, a feature that may play a role in their functional specialization in eliciting care from parents.  

This work just starts to scratch the surface; there is more to be done to fully understand the evolution of vocalization. Future work by Team Squeak will explore the role of variation in deer mouse vocalizations (e.g. do differences in the rate or pitch of cries affect maternal response? Are mothers tuned to features of pup calls from their own species? Do fathers respond in the same way as mothers?) as well as identify how differences in specific genes and ultimately neural circuits give rise to differences in the vocal behaviors of deer mice. Stay tuned! 

By Nick Jourjine and Hopi Hoekstra


 Hopi Hoeksta, Hoekstra Lab


(l to r) Sade McFadden, Anna Lindholm, Juan Sanguinetti-Scheck, Nick Jourjine, Maya Woolfolk, John Sabatini, and Hopi Hoekstra

(l to r) Sade McFadden, Anna Lindholm, Juan Sanguinetti-Scheck, Nick Jourjine, Maya Woolfolk, John Sabatini, and Hopi Hoekstra