*2.6. Functions of Vocalization in Ingestive Behavior*

#### 2.6.1. Alimentary Function

The phenomenon of the social transmission of information about food has been studied in rats for some time [158–160]. The results indicated that the transfer occurs via olfactory cues, and the observers rely on smelling the breath of the demonstrator rat, who has had direct contact with the food [161,162]. Since rats are very vocal in social interactions, the question as to whether rats can convey food preferences via ultrasonic vocalizations was still open and was studied first in female rats. The results suggested that ultrasonic vocalizations do not play role in this communication because information conveyed by the demonstrator rat had no significant influence on the food choices of the observers when the rats were devocalized [163].

This experiment, however, could not fully explain the mechanism of the vocal transmission of feeding information in rats, and the possibility of the vocal transmission of food preferences was recently raised again [164]. Some clues may come from studies performed on female mice, showing that the observer mouse emits ultrasonic vocalizations toward the demonstrator mouse that has been recently fed, but these vocalizations are dependent on the motivational state of the observer. Non-deprived animals emitted more calls toward demonstrators that were fed on palatable food, while food-deprived animals vocalized more to mice that were fed on any food regardless of its palatability [165]. These calls

facilitated the proximity of the mice; however, the exact motivation for the emission of these calls and their communicative value need further studies.

These and other experiments justify distinguishing the category of alimentary calls in rats. In an earlier study, the structure of rat vocalizations emitted by pairs of rats (and recorded in pairs) was studied and categorized, and then the categories were assigned to specific behaviors [166]. Three clusters of ultrasonic calls were identified, roughly referring to frequency-modulated 50 kHz calls, flat 50 kHz calls, and 22 kHz calls. It is of interest that the middle cluster, which was equivalent to the flat type, 50 kHz vocalization (with frequency range between 35 and 55 kHz), contained calls that were emitted mostly during feeding and their emission was consistent [166]. The communicative role of flat versus frequency-modulated 50 kHz vocalizations is dissimilar. For example, during experiments with the self-administered playback of 50 kHz vocalizations, rats reliably self-administered frequency-modulated 50 kHz calls with trills but not the flat 50 kHz calls [73].

#### 2.6.2. Food Provisioning Function

In another recent study, pairs of rats were tested in a mutual food-provisioning task [167]. Firstly, it was found that receiver rats emitted 50 kHz ultrasonic vocalizations toward their donor partners, and the donors provided food to the receivers by pulling a tray with a treat toward the partner rat without a reward for themselves. This food delivery was done in a proportional way to the receivers' communication [167]. These results justify distinguishing a category of ultrasonic calls that have a food-provisioning function. Further research, however, is needed to better understand the type and behavioral role of these calls. It is not clear whether rats were only expressing the need and/or requesting food, or advertising sources of food to other rats.

It seems, however, that the 50 kHz vocalizations in this situation have positive signaling value, i.e., they would be associated with appetitive expectation of food, approach, and eating behavior. On the other hand, it has been shown in the past that alarming 22 kHz vocalizations had an opposite effect on eating behavior. When 22 kHz alarm calls signaled the proximity of a predator (cat), eating behavior was inhibited for up to 2 h [168]. When the alarm subsided, rats emerged from their burrows and resumed eating, but eating bouts were shortened and frequently interrupted by careful observing of the environment [169]. Thus, alarm calls discourage eating, so they have the opposite function to the food provisioning one.

#### *2.7. Functions of Vocalization in Defense against External Threat*

#### 2.7.1. Predator Alarming Function

The most known and well-studied function of rat ultrasonic vocalizations is the alarming function. It evolved as one of the fundamental antipredator behaviors [168–171]. The alarm calls are long-duration 22 kHz vocalizations with relatively constant, i.e., unmodulated, sound frequency and are emitted for a prolonged time, call after call, after detection of the predator and for about 30 min after the predator has left [170,171]. The alarm calls are directed to the members of the entire social group (audience effect) and related to the approaching danger. However, the social effect may not be present, i.e., a rat may emit alarm calls when it is isolated from the group (particularly in the laboratory) or when it may not know where other conspecifics are [172]. Alarm calls are emitted from the place of a relative safety (not in the immediate reach of the predator when the fear response appears) and are not directed to the predator [173]. The alarm vocalizations usually cause a freezing response of the recipient rats, or their escape to the burrows. This effect was reproduced in an experimental situation in which rats that were chased by a fast-moving object (as a potential predator) showed an escape response with the emission of 22 kHz vocalizations and freezing episodes [174].

It has been documented that rats are also highly afraid of predator odors and consistently respond to them with defensive behavior [175]. The odors originate from predators' skin and fur, urine, feces, and anal gland secretions [176]. Rats respond to odors of

many predators (e.g., cat, fox, or lion) but the alarm response to the cat's odor is the strongest [177,178]. When rats were placed in a protective tube within the cage of a predator, they reliably emitted 22 kHz alarm calls to the odor of a cat but emitted only a few calls to that of a snake, and no calls to the odor of a ferret or a control, clean cage [179]. Interestingly, rats did not raise an alarm to the odor of ferrets, which are large carnivores and pose a danger to rats. Ferrets, related to polecats and weasels, however, have been domesticated for a very long time, probably for 2000 years [180]. They live with humans and this could cause some changes in their bodily odors, and it may explain why rats did not recognized the ferret's odor as a threat in this study.

Some analyses of alarming vocalizations of many species led initially to the suggestion that alarm calls evolved to be communicated to predators [181], but other observations have not supported this view. Although cats and other large land predators can hear ultrasonic calls of rats, it was postulated that the ultrasonic alarm calls in rats evolved to protect them from birds of prey (a couple of hundred of species of them) that are the most dangerous predators to rodents [182]. This protection against predators could evolve by adaptation or by exaptation, i.e., by use of naturally preexisting ultrasonic sounds produced by narrow airways for disguised communication. Birds of prey cannot hear ultrasonic calls and their usual audibility is between 1 and 4 kHz (with the exception of the tawny owl, hearing up to 20 kHz) [183]. Thus, communication in the ultrasonic range is adaptive and protects rats. When cornered by a cat, rats have the capacity to defend themselves and often do that successfully as it was indirectly confirmed by a recent publication providing evidence that feral cats were ineffective in hunting for urban rats [184]. Rats have very sharp incisors, and their bites leave deep and not-well-healing wounds infected by bacteria carried by rats [185]. However, rats are defenseless against fast moving birds of prey, which can reach a velocity of 52–70 m/s in extreme situations, as measured for the falcon [186], or developed adaptation for a silent flight as that one of owls [187].

Behavioral analysis of the emission of 22 kHz alarm calls and audible squeals that were emitted in dangerous confrontations with predators or large mammals led to the conclusion that ultrasonic 22 kHz calls are directed to other rats and are associated with an audience effect, while audible squeals are emitted as warning calls directly to predators and other large animals (including humans) and do not require the presence of other conspecifics [173].

The alarming function of 22 kHz ultrasonic calls does not serve individual protection but is a form of social anti-predator defense, i.e., emission of these calls warns the entire social group. This social behavior was regarded as a higher order of defense [18,188]. Hearing the alarm, rats will respond (usually by escape and hiding) regardless of whether the individual colony members have detected or not detected the presence of the predator [171]. Once initiated (usually by alpha male), the alarm is maintained and emitted repeatedly by everyone in the colony for a prolonged time as studied in the visible burrow system [171]. At that time, animals reduce their activity or freeze as it was studied in many independent experiments with the replay of 22 kHz calls [178,189,190].

#### 2.7.2. Alarming-Warning Function

Alarming 22 kHz calls are emitted not only in response to predators, but also when rats encounter any other direct danger. Thus, the alarm calls serve as a general danger signal, so they possess an alarming–warning function to other rats. The alarm calls may be emitted in contact with an unfamiliar human [191,192], in response to a sudden noise (startling acoustic stimulus) [193,194], an unpredictable tactile stimulus with a hissing sound (airpuff) [188,195], or an electric foot shock or tail shock [196–201]. Thus, the semiotic value of the 22 kHz calls in these situations is not related to the predator but is sent as a signal of a general but real and present danger, even though the animal might not fully recognize the nature of the danger (e.g., air-puff) and the danger may not necessarily relate to other rats. These alarming–warning calls are like predator alarm calls and are often interpreted

as alarm calls but they differ from them by the circumstances. The alarming–warning category of calls was usually studied in single rats in laboratory cage conditions.

#### 2.7.3. Security Function

It has been also suggested that the emission of 22 kHz vocalizations may serve as a signal of potential danger that is not actually perceived at a given time by the rat. This emission would be initiated by a special motivational system, termed security motivation system, which was well studied in rats [202–204]. It was argued that this system evolved to cope with unpredictable environmental risks, uncertainty, and potential, but not directly observed or detected, dangers [205]. In this situation, the emission of 22 kHz calls would have a security function. Such alarm 22 kHz vocalizations would serve as precautionary signaling of a potential danger, i.e., as an apprehension signal before any danger appeared [205]. The triggering events could be cues originating from the similarity of an environmental situation to the past aversive events, or the lack of stimuli that the animal would expect to detect (e.g., disappearance of a nearby predator), or some weak or new stimuli unknown to the animal.

The emission of calls from security motivation has not been described; however, there are some fragmentary observations suggesting such motivation. We observed such an emission in the laboratory, when a rat sitting quietly and silently in a cage suddenly started emitting 22 kHz calls without any provocation or other stimuli detectable to humans. The category of anticipatory calls has already been demonstrated for vocalizations expressing a positive emotional state associated with drugs of abuse [206]. Security motivation signaling requires further systematic studies.

All the functions described in the subsections above are associated with significant endocrine and autonomic changes, such as release of ACTH, changes in blood pressure, heart rate, body temperature, and respiration rate and confirm the stressful and emotional nature of situations associated with the emissions of ultrasonic vocalizations [207–210]. Although initial recordings of heart rate during the playback of ultrasonic calls did not detect changes in heart rate [211], more detailed and frequent sampling of heart rate detected changes caused by the playback. It is interesting that receivers of the ultrasonic calls develop relevant emotional arousal with autonomic changes [212]. Playback of 22 kHz vocalizations decreased heart rate in the receiver rats, while the playback of 50 kHz calls increased the rats' heart rate. These effects were stronger in singly housed rats as compared to pair-housed rats [212].
