Vocal communication between individuals is vital to navigating social situations. Often when we think about vocal communication, we think about speech in humans or song in birds which is used to convey information from one individual to another. However, recently rodents have emerged as an exciting new mammalian experimental model of vocal communication.
While rodents are well known for their great sense of smell and use of olfactory communication it turns out that they also use a wide variety of vocalizations to convey information amongst individuals. You may have heard about singing mice, but the majority of rodent vocalizations are inaudible to humans as they are produced in the ultrasonic range. These ultrasonic vocalizations, or USVs, are generated at extremely high frequencies, from 20 – 100 kHz compared to a maximum of 4 kHz in humans! Despite knowing that rodents communicate vocally, we know very little about what these vocalizations mean.
To this point, work on rodent USVs has mostly focused primarily on mating and emotion. In rats and mice, these USVs help to attract mates, and rats that don’t call or produce strange calls will not be attractive to females. Rodent calls are also thought to display emotion with lower frequency calls, around 22 kHz, displaying distress while higher frequency calls, above 50 kHz, generally indicate happiness. However, can these calls represent other information than emotions? Are these calls used in contexts other than mating and courtship? Our answer is yes.
You can listen to the calls here: https://soundcloud.com/pcalusv
Our current study investigating this research represents work initiated by Dr. Nathan Rieger who recently obtained his Ph.D. from the University of Wisconsin – Madison. The research team was interested in understanding what calls are used to display territoriality to an intruder and potentially warn of future aggression. The California mouse was the ideal species for this because they use ultrasonic vocalizations, exhibit strict monogamy, both parents raise the young and both sexes are highly territorial. During all of these social interactions, they are vocalizing to each other in an auditory range that we cannot hear. Importantly, both males and females will vocalize and defend their territories from intruders.
We focused on two call types produced by California mice. The first is called a sustained vocalization. When slowed down significantly, these sound almost like a whale call. The calls are relatively long for a rodent ultrasonic vocalization and can be produced either in isolation or as a number of syllables within a chain. Sustained vocalizations do not fluctuate much in their frequency but their length and the number of syllables produced can vary widely based on social context.
The second type of call we focused on are called barks. Barks earned their name because when they are slowed down to be fully in the human audible range they sound just like a ferocious bark! (Seriously, check out our SoundCloud below.) Barks are a special case of USV as they actually start in the human audible range, become ultrasonic, and then return to being audible again in quick succession. The function of barks before this study was not known but we predicted that they would be involved in agonistic interactions because these calls are often recorded in parallel with the display of other agonistic behaviors that occur when newly introduced males and females first encounter each other. Finally, despite the fact that a new category of USVs did not relate to territorial behavior, we discovered a new call, sweep-phrases. These calls sound like birdsong when slowed down into the vocal range and display a level of complexity not previously seen in rodents.
To test how barks and sustained vocalizations were used by male and female California mice we created a paradigm where individual mice were challenged in their territory by an unknown mouse of their same-sex. We worked with pair-bonded mice that are more likely to protect their territories and, thus, we believed would be more likely to advertise and defend their territories. We then maintained either the male or female member of the pair in the territory while removing their partner and introduced a novel individual into the territory on the opposite side of a Plexiglas divider. This divider allowed us to examine what kind of USVs would be made towards an intruder that the focal mouse could see and smell but not approach, indicative of advertising a territory. We then removed the divider and allowed the two mice to interact and recorded their vocalizations. After this interaction, the divider was replaced and we recorded vocalizations that occurred after the resolution of the interaction.
We found that while males and females did not differ in how they physically protected their territory – they both were able to successfully repel the intruder – there were sex differences in how they vocalized during these interactions, and their vocalizations changed based on whether or not they could physically interact with the intruder. First, we saw that paired males, but not females or intruders, produced sustained vocalizations prior to physical interaction when the divider was in place. This likely means that males are more likely to advertise their territories than females are. However, once physical interactions started, females produced more barks than males. These barks, interestingly, were produced by intruders. We suggest that these calls could be used to signal submission to the resident mouse. Finally, after the physical interaction was terminated, almost all vocalizations stopped which could mean that the resolution of the interaction renders communication between individuals no longer necessary.
We also found that resident male California mice shortened their sustained vocalizations prior to a physical interaction and that shorter calls predicted a more rapid approach to the intruder. Shorter chains of sustained vocalizations also predicted a more rapid approach. As such, changes in the sustained vocalizations could warn intruders that the male is ready to defend its territory against the intruder, likely to induce the intruder to retreat. We also found that calls were shortest while physical interaction was occurring, much like humans might switch to shorter words during a heated exchange, California mice switch to shorter calls while interacting with unwanted guests.
These results show that California mice use vocalizations to navigate complex social situations like an intruder entering their territory. Moreover, it shows that California mice potentially can modify their vocalizations to convey information about their future behavior to another individual, warning them that they are prepared to defend their territory. This shows a level of complexity in rodent vocal behavior that has not been previously explored and potentially provides a new system by which we can understand the role of communication in rodent social behavior.
In summary, yes, California mice use ultrasonic vocalizations in contexts other than mating scenarios. Females, not just males, will also energetically defend their territories. Notably, USVs play an integral part in this defense. There is information content in at least male USVs that could be used by other individuals of the same species to gauge their willingness to defend a territory. Have we ruled out that these rodent calls are only expressing emotion based solely on frequency? The answer to this would be yes, because of the information content in length of calls and altered number of calls in bouts. More studies, however, will be needed to fully test whether the USVs convey more than emotional state. At the very least, we demonstrate a level of complexity in rodent vocal behavior that has not been previously explored and builds on an experimental model that was identified by others through field research.
These findings are described in the article entitled The function of ultrasonic vocalizations during territorial defence by pair-bonded male and female California mice, recently published in the journal Animal Behaviour. This work was conducted by Nathaniel S. Rieger and Catherine A. Marler from the University of Wisconsin-Madison.