Communication is ubiquitous in animals and involves one individual influencing the behavior of another individual through the use of signals or cues (see the “Communication, Cues, and Signal” entry for more information). Animals often communicate when they detect a potential predator by producing alarm signals. Although such signals can be produced via different modalities, many species rely on vocal signals – alarm calls (Bradbury and Vehrencamp 2011). Alarm calls function to alert others of the potential danger related to a perceived predator. In order for such communication to be adaptive, the benefits associated with alarm calling in the presence of a predator must be greater than the costs.
Both the physical structures of alarm calls and the motivations behind their production are important to understanding signal function (Bradbury and Vehrencamp 2011). In birds and mammals, and possibly other groups, vocal signals that are low-frequency and harsh are often produced by individuals behaving aggressively. Low-frequency sounds are easier to locate; animals often use calls with this structure to mob or indicate distress. Signals that are higher in frequency and that possess relatively little frequency modulation are often produced by fearful signalers. These signals are more difficult to localize and likely function as alarm calls meant primarily for receivers of the same species as the signaler. Receivers often freeze or flee when high-frequency alarm calls are produced.
When considering the adaptive functions of alarm calls, the costs and benefits of producing an alarm signal must be determined. For alarm calling to be adaptive, the benefits of signaling must outweigh the costs over time. There are several hypotheses to explain the adaptive benefits of alarm signaling (Caro 2005; Wheeler 2008). Under the kin selection hypothesis, an alarm signaler will risk making an alarm call and attracting the attention of the predator because the signal may increase the sender’s indirect fitness by alerting and potentially saving related family members. Similarly, the parental care hypothesis proposes that the signaler will risk alarm calling because the signal may increase their direct fitness by alerting their own offspring. Under the selfish herd hypothesis, a sender produces an alarm call to attract other individuals, thus increasing the number of individuals present and thereby reducing the probability of predation for the signaler. Similarly, the mobbing recruitment hypothesis predicts that a signaler risks making an alarm call because it will attract others to the area to harass the predator and increase the chances that the predator leaves. The predator confusion hypothesis states that the signaler risks making an alarm call because it may cause a sudden burst of movement from others as they flee, which confuses the predator and reduces the threat. Lastly, with the pursuit deterrence hypothesis, a signaler risks sending an alarm call because the call may cause the predator to give up the hunt after it has been spotted by its prey. These hypotheses are not mutually exclusive; for example, when testing the adaptive function of alarm calling in wild capuchins, Cebus apella nigritus, Wheeler (2008) found that in the presence of a predatory viper, the kin selection, parental care, and mobbing recruitment hypotheses were all supported. Capuchins, therefore, alarm call to recruit others to mob the predator but also to communicate to their offspring and related family members. The functions of alarm calls may thus differ depending on the composition of the group and the predator threat level.
Semantic Alarm Calling
Some individuals can encode specific information in their alarm calls about the type of predator, the predator threat level, or both (Gill and Bierema 2013). Semantic, or referential, alarm calls are signals that refer to specific predator types or predator behavior patterns (Seyfarth et al. 1980). In a seminal experiment, Seyfarth et al. (1980) used a playback method to test different alarm calls produced by vervet monkeys, Cercopithecus aethiops, in response to three different predators: leopards, eagles, and snakes. These varying alarm calls were considered referential because they were each produced by signalers in specific contexts (i.e., each was produced in relation to the specific predator detected), and the behavior of the receivers differed depending on the type of predator that elicited the alarm calls. For example, when vervet monkeys heard playbacks of leopard alarm calls, the monkeys more often ran high up into a tree, which is the adaptive behavioral response to this type of predator. Vervets produced different behavioral responses to playbacks of the eagle and snake alarm calls, revealing specificity in receiver response.
Other species may encode information into their alarm calls that indicates the threat level of the potential predator, rather than the predator type. For example, Siberian jays (Perisoreus infaustus) produce alarm calls when predatory hawks are detected, but the specific structure of the calls is related to whether the predator is perched (lower-risk) or attacking (high-risk; Griesser 2008). Chickadee, tit, and titmouse species also vary the number of specific note types in their “chick-a-dee” calls in relation to the level of risk associated with different sizes of predators they have detected.
Deceptive Alarm Calling
Whereas it is typically adaptive for signalers to produce alarm calls when a predator has been detected, and for receivers to react to alarm calls appropriately, some individuals use alarm calls in non-predator contexts in deceptive ways. These false alarm calls are seen in several species. For example, great tits, Parus major, alarm call to deceive competitors of the same species as the signaler for food, even when there is no predator threat (Møller 1988). These deceptive calls cause surrounding individuals to be temporarily distracted and/or flee, thus allowing the deceptive signaler better access to limited food resources.
Deceptive alarm calling can occur between members of different species. For example, the fork-tailed drongo, Dicrurus adsimilis, is able to mimic the alarm calls of meerkats, Suricata suricatta, and can deceive meerkats by producing the meerkats’ own species-specific alarm call (Flower 2010). Meerkats eavesdrop on the alarm calls of drongos and thus benefit from their species-specific alarm calls. Meerkats that react to a drongo’s mimicked meerkat alarm incur the cost of abandoning found food. However, because drongos also regularly produce honest alarm calls, the meerkats benefit from using a “better safe than sorry” strategy when drongos alarm call (Flower 2010). A general view of deceptive alarm calling is that such calling persists in many systems because the costs of responding to false alarm calls (when a predator is not present) in a food context are much smaller than the costs of failing to respond appropriately to true alarm calls (when a predator is present).
Alarm calls are one class of vocal signaling related to anti-predator behavior (Caro 2005). Other classes of signal used in anti-predator contexts include predator deterrence, mobbing, and distress signals. Predator deterrence signals function minimally to communicate to the predator that it has been detected, and may also provide information regarding the signaler’s condition or evasion ability. Mobbing calls function to attract members of the same and different species as the signaler to gather, harass, and deter the predator to drive it from the area. Individuals may also produce distress calls when they are physically captured or constrained by a predator, potentially to cause the predator to release the distressed individual.
Alarm calls represent an important class of animal signals in that they are strongly related to individual fitness and also because many of them seem to involve complicated cognitive processing on the part of signalers or receivers or both. For example, alarm calls falsely produced in contexts of contested resources (like food or mates) often represent cases of tactical deception and suggest signaler awareness of perceptions or knowledge of receivers. In some species, signalers vary production of alarm calls in sophisticated ways to communicate about not just predator presence but also predator behavior. Additionally, some of the most structurally complex signals in animal communication function in anti-predator behavior. One example is the already mentioned “chick-a-dee” call of black-capped chickadees and related species. Another example involves the syntax-like combination of different vocalizations in putty-nosed monkeys (Cercopithecus nictitans) to signal increased predator threat (Arnold and Zuberbühler 2004).
Taken together, alarm calls are an important type of vocal communication that signal about potential predator threats. While these vocalizations may draw the attention of a predator, species that utilize alarm calls do so in an adaptive way, with benefits generally outweighing the costs of signaling. There are several hypothesized functions of alarm calls, including to alert others, to recruit others, and to confuse or deter the predator. Further research can highlight the adaptive value of more nuanced aspects of alarm signaling, such as distinctive alarm calls made by parents for their offspring compared to other adults in the immediate environment (e.g., Suzuki 2011).