On the Meaning of Alarm Calls: A Review of Functional Reference in Avian Alarm Calling

A long‐standing question in animal communication is whether signals reveal intrinsic properties of the signaller or extrinsic properties of its environment. Alarm calls, one of the most conspicuous components of antipredator behaviour, intuitively would appear to reflect internal states of the signaller. Pioneering research in primates and fowl, however, demonstrated that signallers may produce unique alarm calls during encounters with different types of predators, suggesting that signallers through selective production of alarm calls provide to conspecific receivers information about predators in the environment. In this article, we review evidence for such ‘functional reference’ in the alarm calls of birds based on explicit tests of two criteria proposed in Macedonia & Evans’ (Ethology 93, 1993, 177) influential conceptual framework: (1) that unique alarm calls are given to specific predator categories, and (2) that alarm calls isolated from contextual information elicit antipredator responses from receivers similar to those produced during actual predator encounters. Despite the importance of research on birds in development of the conceptual framework and the ubiquity of alarm calls in birds, evidence for functionally referential alarm calls in this clade is limited to six species. In these species, alarm calls are associated with the type of predator encountered as well as variation in hunting behaviour; with defence of reproductive effort in addition to predators of adults; with age‐related changes in predation risk; and with strong fitness benefits. Our review likely underestimates the occurrence of functional reference in avian alarm calls, as incomplete application and testing of the conceptual framework has limited our understanding. Throughout, therefore, we suggest avian taxa for future studies, as well as additional questions and experimental approaches that would strengthen our understanding of the meaning of functional reference in avian alarm calls.     

Semantic communication in birds: evidence from field research over the past two decades

What do animal signals mean? This is a central question in studies on animal communication. Research into the semantics of animal signals began in 1980, with evidence that alarm calls of a non-human primate designated predators as external referents. These studies have challenged the historical assumption that such referential signaling is a unique feature of human language and produced a paradigm shift in animal communication research. Over the past two decades, an increasing number of field studies have revealed similar complexity in anti-predator communication of birds. The acoustic structures of avian alarm calls show a high degree of variation in pitch, duration, shape, and repetition rate. In addition to such distinct and graded variations, several birds combine discrete types of notes or calls into higher complex sequences. These variations in alarm calls are typically associated with the predator’s attributes, such as predator type and distance, and receivers respond to them with appropriate anti-predator behaviors. Although alarm calls of several bird species, as well as those of monkeys, appear to denote predator attributes, almost nothing is known about the cognitive processes that underlie the production and perception of these signals. In this review, I explore the existing evidence for referential signaling in birds and highlight the importance of the cognitive approach to animal communication research. I hope this review will promote further investigations of alarm-calling behavior in birds and will help enhance our understanding of the ecology and evolution of semantic communication.     

Motivation before meaning: motivational information encoded in meerkat alarm calls develops earlier than referential information

In contrast to historical assumptions about the affective nature of animal vocalizations, it is now clear that many vertebrates are capable of producing specific alarm calls in response to different predators, calls that provide information that goes beyond the motivational state of a caller. However, although these calls function referentially, it does not mean that they are devoid of motivational content. Studies on meerkats (Suricata suricatta) directly support this conclusion. The acoustic structure of their alarm calls simultaneously encodes information that is both motivational (level of urgency) and referential (predator specific). In this study, we investigated whether alarm calls of young meerkats undergo developmental modification and whether the motivational or the referential aspect of calls changes more over time. We found that, based on their acoustic structure, calls of young showed a high correct assignment to low- and high-urgency contexts but, in contrast to adults, low assignment to specific predator types. However, the discrimination among predator types was better in high-urgency than in low-urgency contexts. Our results suggest that acoustic features related to level of urgency are expressed earlier than those related to predator-specific information and may support the idea that referential calls evolve from motivational signals.     

The acoustic structure of suricates'' alarm calls varies with predator type and the level of response urgency.

The variation in the acoustic structure of alarm calls appears to convey information about the level of response urgency in some species, while in others it seems to denote the type of predator. While theoretical models and studies on species with functionally referential calls have emphasized that any animal signal considered to have an external referent also includes motivational content, to our knowledge, no empirical study has been able to show this. In this paper, I present an example of a graded alarm call system that combines referential information and also information on the level of urgency. Acoustically different alarm calls in the social mongoose Suricata suricatta are given in response to different predator types, but their call structure also varies depending on the level of urgency. Low urgency calls tend to be harmonic across all predator types, while high urgency calls are noisier. There was less evidence for consistency in the acoustic parameters assigned to particular predator types across different levels of urgency. This suggests that, while suricates convey information about the level of urgency along a general rule, the referential information about each category of predator type is not encoded in an obvious way.     

Context-specific calls in wild chimpanzees, Pan troglodytes verus: analysis of barks

Context-specific calls, which have a distinct acoustic structure and are selectively produced in specific contexts, are a prerequisite for calls that function referentially. Functionally referential calls, which convey information to conspecifics about objects and events in the external world, have been found in a number of species, notably primates. Evidence of context-specific calls in apes, however, is largely absent. We analysed whether the barks of wild male chimpanzees in the Ta? Forest, Côte d'Ivoire, are context specific. We examined the acoustic structure of barks, and other calls produced in association with barks, in six contexts, using discriminant function analysis. Chimpanzees produced context-specific signals in two ways. First, they produced two acoustically graded bark subtypes, in hunt and snake contexts, respectively. Second, they produced context-specific signal combinations of barks with acoustically different call types or drums. These signal combinations increased specificity levels in three of the six contexts to over 90%, a level similar to the classic vervet monkey, Cercophithecus aethiops, predator alarm calls. Furthermore, specific chimpanzee signals were produced in contexts other than alarm, such as travel and hunting, where the potential benefits of evolving specific calls are less obvious. These signals may convey specific context information to listeners, and thus function referentially; however, to confirm this, analyses of listeners' responses are required. The results show that two strategies for producing context-specific signals seem to have evolved in a species other than humans: chimpanzees produce context-specific bark subtypes and context-specific signal combinations. Copyright 2003 Published by Elsevier Science Ltd on behalf of The Association for the Study of Animal Behaviour.      

Eavesdropping on heterospecific alarm calls: from mechanisms to consequences

Animals often gather information from other species by eavesdropping on signals intended for others. We review the extent, benefits, mechanisms, and ecological and evolutionary consequences of eavesdropping on other species' alarm calls. Eavesdropping has been shown experimentally in about 70 vertebrate species, and can entail closely or distantly related species. The benefits of eavesdropping include prompting immediate anti‐predator responses, indirect enhancement of foraging or changed habitat use, and learning about predators. Eavesdropping on heterospecifics can provide more eyes looking for danger, complementary information to that from conspecifics, and potentially information at reduced cost. The response to heterospecific calls can be unlearned or learned. Unlearned responses occur when heterospecific calls have acoustic features similar to that used to recognize conspecific calls, or acoustic properties such as harsh sounds that prompt attention and may allow recognition or facilitate learning. Learning to recognize heterospecific alarm calls is probably essential to allow recognition of the diversity of alarm calls, but the evidence is largely indirect. The value of eavesdropping on different species is affected by problems of signal interception and the relevance of heterospecific alarm calls to the listener. These constraints on eavesdropping will affect how information flows among species and thus affect community function. Some species are ‘keystone’ information producers, while others largely seek information, and these differences probably affect the formation and function of mixed‐species groups. Eavesdroppers might also integrate alarm calls from multiple species to extract relevant and reliable information. Eavesdropping appears to set the stage for the evolution of interspecific deception and communication, and potentially affects communication within species. Overall, we now know that eavesdropping on heterospecific alarm calls is an important source of information for many species across the globe, and there are ample opportunities for research on mechanisms, fitness consequences and implications for community function and signalling evolution.     

The evolution of functionally referential meaning in a structured world

Animal communication systems serve to transfer both motivational information--about the intentions or emotional state of the signaler--and referential information--about external objects. Although most animal calls seem to deal primarily with motivational information, those with a substantial referential component are particularly interesting because they invite comparison with words in human language. We present a game-theoretic model of the evolution of communication in a "structured world", where some situations may be more similar to one another than others, and therefore require similar responses. We find that breaking the symmetry in this way creates the possibility for a diverse array of evolutionarily stable communication systems. When the number of signals is limited, as in alarm calling, the system tends to evolve to group together situations which require similar responses. We use this observation to make some predictions about the situations in which primarily motivational or referential communication systems will evolve.     

The versatility of graded acoustic measures in classification of predation threats by the tufted titmouse Baeolophus bicolor: exploring a mixed framework for threat communication

Many mammal and bird species respond to predator encounters with alarm vocalizations that generate risk‐appropriate responses in listeners. Two conceptual frameworks are typically applied to the information encoded in alarm calls and to associated anti‐predator behaviors. ‘Functionally referential’ alarm systems encode nominal classes or categories of risk in distinct call types that refer to distinct predation‐risk situations. ‘Risk‐based’ alarms encode graded or ranked threat‐levels by varying the production patterns of the same call types as the urgency of predation threat changes. Recent work suggests that viewing alarm‐response interactions as either referential or risk‐based may oversimplify how animals use information in decision‐making. Specifically, we explore whether graded alarm cues may be useful in classifying risks, supporting a referential decision‐making framework. We presented predator (hawk, owl, cat, snake) and control treatments to captive adult tufted titmice Baeolophus bicolor and recorded their vocalizations, which included ‘chick‐a‐dee’ mobbing calls (composed of chick and D notes), ‘seet’ notes, two types of contact notes (‘chip’, ‘chink’), and song. No single call type was uniquely associated with any treatment and the majority of acoustic measures varied significantly among treatments (46 of 60). The strongest models (ANOVA and classification tree analysis) grouped hawk with cat and owl, and control with snake, and were based on the number or proportion of a) chick and D notes per chick‐a‐dee call, b) chip versus chink notes produced following treatment exposure, and c) the frequency metrics of other note types. We conclude that (1) the predation‐threat information available in complex titmouse alarm calls was largely encoded in graded acoustic measures that were (2) numerous and variable across treatments and (3) could be used singly or in combinations for either ranking or classification of threats. We call attention to the potential use of mixed threat identification strategies, where risk‐based signal information may be used in referential decision‐making contexts.     

Marmoset (Callithrix geoffroyi) Food-Associated Calls are Functionally Referential

Signals that are functionally referential provide listeners with information about a signaler's external environment, such as the presence of a nearby predator. Just as alarm calls may signal the presence of a predator, food-associated calls may signal the presence of food. To date, however, the only conclusive evidence that conspecifics perceive information about food from food-associated calls comes from a single species, the domestic chicken. Geoffroy's marmosets, small, arboreal New World primates, often emit food-associated calls when finding or consuming food. We presented playbacks of food calls and control sounds to two groups of naturalistically housed marmosets and observed the frequency of food-related behaviors (FRBs) during the 20-min post-playback period. Relative to the control playbacks, food call playbacks elicited an increase in the frequency of two FRBs (foraging and feeding), lasting throughout the post-playback observations. Moreover, the effect of food call playbacks was robust, occurring without regard to the type of food eliciting the food calls, the rate of food calling, or whether the signaler was a member of the receiver's group. To our knowledge, this research is the first to demonstrate that the food-associated calls of a primate species, like the food-associated calls of domestic chickens, meet the perception criterion for functionally referential communication. The results are discussed in light of the affective and cognitive mechanisms that may underlie the responses of receivers to hearing food-associated calls given by marmoset conspecifics.     

Learning and signal copying facilitate communication among bird species

Signals relevant to different sets of receivers in different contexts create a conflict for signal design. A classic example is vocal alarm signals, often used both during intraspecific and interspecific interactions. How can signals alert individuals from a variety of other species in some contexts, while also maintaining efficient communication among conspecifics? We studied heterospecific responses to avian alarm signals that drive the formation of anti-predator groups but are also used during intraspecific interactions. In three species-rich communities in the western Himalayas, alarm signals vary drastically across species. We show that, independently of differences in their calls, birds respond strongly to the alarm signals of other species with which they co-occur and much more weakly to those of species with which they do not co-occur. These results suggest that previous exposure and learning maintain heterospecific responses in the face of widespread signal divergence. At an area where only two species regularly interact, one species'' calls incorporate the call of the other. We demonstrate experimentally that signal copying allows strong responses even without previous exposure and suggest that such hybrid calls may be especially favoured when pairwise interactions between species are strong.     

Functionally referential signals: A promising paradigm whose time has passed

Finding the evolutionary origins of human language in the communication systems of our closest living relatives has, for the last several decades, been a major goal of many in the field of animal communication generally and primate communication specifically. 1 - 4 The so‐called “functionally referential” signals have long been considered promising in this regard, with apparent parallels with the semantic communication that characterizes language. The once‐prominent idea that functionally referential signals are word‐like, in that they are arbitrary sounds that refer to phenomena external to the caller, has largely been abandoned. 5 However, the idea that these signals may offer the strongest link between primate communication and human language remains widespread, primarily due to the fact the behavior of receivers indicates that such signals enable them to make very specific inferences about their physical or social environment. Here we review the concept of functional reference and discuss modern perspectives that indicate that, although the sophistication of receivers provides some continuity between nonhuman primate and human cognition, this continuity is not unique to functionally referential signals. In fact, because functionally referential signals are, by definition, produced only in specific contexts, receivers are less dependent on the integration of contextual cues with signal features to determine an appropriate response. The processing of functionally referential signals is therefore likely to entail simpler cognitive operations than does that of less context‐specific signals. While studies of functional reference have been important in highlighting the relatively sophisticated processes that underlie receiver behavior, we believe that the continued focus on context‐specific calls detracts from the potentially more complex processes underlying responses to more unspecific calls. In this sense, we argue that the concept of functional reference, while historically important for the field, has outlived its usefulness and become a red herring in the pursuit of the links between primate communication and human language. © 2012 Wiley Periodicals, Inc.     

Alarm Calls of Marmosets (Callithrix geoffroyi) to Snakes and Perched Raptors

Communication about the presence of predators is an important benefit of group living. Critical information about the nature of danger can be conveyed through referential alarm calls. Raptors pose a significant predatory threat to callitrichid species. Unlike a raptor in flight, a perched raptor cannot attack suddenly at great speed, and it can be monitored from a safe distance. In this sense a perched bird may pose a threat more similar to that of a terrestrial predator such as a snake. Here we compare predatory contexts by addressing these two questions: 1) Do marmosets produce acoustically distinct alarm calls to snake models and perched raptor models? 2) Do the visual responses of the marmosets to the playbacks of perched raptor–elicited calls differ from those given to the playbacks of calls given in response to snakes? We recorded alarm calls from two groups of outdoor-housed Geoffroy’s marmosets (Callithrix geoffroyi) in response to predator models. Later, we played back stimuli created from these recordings to the marmosets and scored their gaze direction. Results show that calls given to models of perched raptors are acoustically distinct from those given to models of snakes. Further, the relative number of upward to downward looks while listening to the playbacks of perched raptor–elicited calls was significantly greater than it was for snake-elicited calls. Reactions to airborne raptors are known to elicit freezing or rapid flight, neither of which occurred in response to our playbacks. Our data suggest a greater complexity in the alarm call repertoire of marmosets than previously demonstrated.     

Perceptual specificity in the alarm calls of Gunnison's prairie dogs

Gunnison's prairie dogs have a complex alarm communication system. We show that the escape responses of prairie dogs to naturally occurring live predators differed depending upon the species of predator. We also show that playbacks of alarm calls that were elicited originally by the live predators produced the same escape responses as the live predators themselves. The escape responses fell into two qualitatively different categories: running to the burrow and diving inside for hawks and humans, and standing upright outside the burrow for coyotes and dogs. Within these two categories there were differences in response. For hawks, only the prairie dogs that were in the direct flight path of a stooping red-tailed hawk ran to their burrows and dove inside, while for humans and human alarm call playbacks there was a colony-wide running to the burrows and diving inside. For coyotes and coyote alarm call playbacks there was a colony-wide running to the burrows and standing alert at the burrow rims, while for domestic dogs and playbacks of alarm calls for domestic dogs the prairie dogs assumed an alert posture wherever they were feeding, but did not run to their burrows. These responses to both the live predators and to predator-elicited alarm calls suggest that the alarm calls of Gunnison's prairie dogs contain meaningful referential information about the categories of predators that approach a colony of prairie dogs.     

Heterospecific recognition of referential alarm calls in two species of lemur

ABSTRACT

Alarm vocalizations are a common feature of the mammalian antipredator response. The meaning and function of these calls vary between species, with some species using calls to reference-specific categories of predators. Species can also use more than just the calls of conspecifics to detect threat, ‘eavesdropping’ on other species’ signalling to avoid predation. However, the evidence to date for both referential signalling and eavesdropping within primates is limited. We investigated two sympatric populations of wild lemur, the Coquerel’s sifaka Propithecus coquereli and the common brown lemur Eulemur fulvus, presenting them with playbacks of predator calls, conspecific alarm calls and heterospecific lemur alarm calls, and recorded their behavioural responses following the playbacks. Results suggest that the Coquerel’s sifaka may have functionally referential alarm calls with high specificity for aerial predators, but there was no evidence for any referential nature of the other call investigated. Brown lemurs appear to have a mixed alarm system, with one call being specific with respect to aerial predators. The other call investigated appeared to reference terrestrial predators. However, it was also used in other contexts, so does not meet the criteria for functional reference. Both species showed evidence for heterospecific alarm call recognition, with both the Coquerel’s sifaka and the brown lemurs responding appropriately to heterospecific aerial alarm calls.     

Ontogeny of conspecific and heterospecific alarm call recognition in wild Verreaux's sifakas (Propithecus verreauxi verreauxi)

The production of vocalizations in nonhuman primates is predominantly innate, whereas learning influences the usage and comprehension of vocalizations. In this study, I examined the development of alarm call recognition in free-ranging infant Verreaux's sifakas. Specifically, I investigated their ability to recognize conspecific alarm calls as well as those of sympatric redfronted lemurs (Eulemur fulvus rufus) in Kirindy forest, western Madagascar. Both species have functionally referential alarm calls for aerial predators and give general alarm calls for both aerial and general predators and also other kinds of threats, such as intergroup encounters with conspecifics. I conducted playback experiments with members of two birth cohorts (nine and ten individuals) to determine the age at which infant Verreaux's sifakas discriminate between conspecific alarm calls, heterospecific alarm calls, and non-alarm vocalizations (parrot song). Most 3-4 months old infants fled toward adults after hearing any playback stimuli, whereas 4-5-month-old infants did so only after presentation of alarm calls. Moreover, all infants of these age classes showed a longer latency to flee after the parrot song indicating their emerging ability to discriminate between alarm calls and non-alarm stimuli. At an age of about 6 months, infants switched from fleeing toward adults to performing adult-like escape responses after presentation of conspecific and heterospecific alarm calls. Thus, the ability to discriminate between alarm from non-alarm stimuli precedes the appearance of adult-like responses. The transition to adult-like escape behavior was coincident with the physical independence of infants from their mothers.     

Experimental study of alarm calls of the oriental tit (Parus minor) toward different predators and reactions they induce in nestlings

Anti-predatory strategies of birds are diverse and may include predator-specific alarm calls. For example, oriental tit (Parus minor) parents can distinguish snakes from other predators and produce snake-specific referential vocalizations ("jar" call) when a snake poses a threat to their nest. The “jar” call has a very specific function to induce fledging of nestlings close to fledging age. This reaction ensures nestlings' survival in natural encounters with snakes that are capable of entering nest cavities and kill entire broods. Sciurid rodents, like chipmunks, may pose a similar threat to cavity-nesting birds. We explored the hypothesis that parents use the fledging-inducing alarm vocalizations in this situation, because chipmunks, like snakes, can kill the brood upon entering the nest cavity. We compared alarm calls of parents toward two predators (chipmunk and snake) who pose a similar threat to the nestlings in a nest cavity, and toward an avian predator (Eurasian jay) who cannot enter nest cavities and poses no threat to the nestlings in a nest. Our results show that the vocal responses of oriental tits were different among the three predators. This suggests that the acoustic properties of vocal responses to predators are different between predators of a similar hunting strategy (nest-cavity entering). The playback of recorded vocal responses of parents to chipmunks did not trigger the fledging of old nestlings, whereas the vocalizations toward a snake did, as shown by earlier studies. Our study suggests that the vocal response of parents does not carry information about the ability of predators to enter the nest cavity and confirms the special status of alarm calls triggered by snakes.     

Monotony and the Information Content of Richardson's Ground Squirrel (Spermophilus richardsonii) Repeated calls: Tonic Communication or Signal Certainty?

The repetition of elements within an alarm signal is commonly thought to ensure that receivers have detected that signal, or to promote residual vigilance in light of the dangerous circumstances prompting the signal's initial production (tonic communication). Beyond alerting others and maintaining that state of alertness, however, repetitive signal elements may be parsed so as to encode information about the nature of potential threats. To determine how call length and variation in intersyllable latency might prove informative in the repetitive alarm vocalizations of Richardson's ground squirrels (Spermophilus richardsonii), we conducted a field‐based playback experiment quantifying antipredator responses to manipulated alarm calls. Free‐living juvenile squirrels were exposed to playbacks of 12 syllable (long) and six syllable (short) calls with a constant (monotonous) or changing (variable) call rate. The length of calls had no significant effect on the behaviour of call recipients during and immediately after call production; however, call recipients showed greater vigilance after the playback of monotonous calls than after variable calls. The absence of a call length effect is not consistent with tonic communication in the strict sense; rather, enhanced responsiveness to monotonous relative to variable calls suggests that multiple syllables, and the emergent patterns of intersyllable latency, communicate information about response urgency or the distance to a predatory threat. Only monotonous calls convey those aspects with any certainty on the part of the signaller and hence are selectively attended to by receivers.     

What is Communicated in the Antipredator Calls of Lemurs: Evidence from Playback Experiments with Ringtailed and Ruffed Lemurs

Two hypotheses of signal specificity in antipredator calls (“referential signalling” and “response urgency”) are discussed in light of prior research on ground squirrels and vervet monkeys. These hypotheses then are examined with data on responses of semi-captive ringtailed and ruffed lemurs to antipredator call playbacks. Although the responses of ringtailed lemurs support a referential-signalling interpretation of their antipredator calls, those of ruffed lemurs do not conform well to either hypothesis. Rather, ruffed lemur antipredator calls seem best viewed as “affective” signals that may only reflect underlying emotional/motivational states.     

Mule deer (Odocoileus hemionus) respond to yellow‐bellied marmot (Marmota flaviventris) alarm calls

Individuals may obtain valuable information about the presence of predators by listening to heterospecific alarm signals. Most playback studies have demonstrated that similarly sized and taxonomically related species may respond to the calls of each other, but less work has been carried out to define these factors influence responsiveness to alarm signals. In theory, individuals should respond to calls from any species that provide information about the presence of important predators, regardless of body size or taxonomic relationship. However, size is often associated with vulnerability. Coyotes (Canis latrans) in the Rocky Mountains prey upon both mule deer (Odocoileus hemionus) and yellow‐bellied marmots (Marmota flaviventris), which differ considerably in size, alarm vocalizations, and antipredator behavior. We conducted a playback experiment to see whether deer discriminated between marmot alarm calls and the non‐alarm song of a common sympatric bird. We found that deer increased vigilance significantly more after hearing broadcast marmot alarm calls compared with the bird song. Interestingly, deer that were studied within 0.5 km of homes showed significantly greater discrimination than those studied farther from humans. Our results suggest relative size differences do not prevent interspecific communication and that common predators should generally drive the evolution of the ability to learn to respond to meaningful risk cues. As long as two species share a predator, it should benefit the other to respond to its alarm calls.     

Semantic Differences in Sifaka (Propithecus verreauxi) Alarm Calls: A Reflection of Genetic or Cultural Variants?

In this study, we compared the usage of alarm calls and anti‐predator strategies between a captive and a wild lemur population. The wild lemur population was studied earlier in Western Madagascar ( Fichtel & Kappeler 2002 ). The captive population was studied in outdoor enclosures of the Duke University Primate Center. Alarm calls and anti‐predator behavior were elicited by conducting experiments with both vocal and visual dummies. We scored the subjects’ immediate behavioral responses, including alarm calls, from video recordings made during the experiments. In principle, both populations have a mixed alarm call system with functionally referential alarm calls for aerial predators and general alarm calls for terrestrial and aerial predators and for situations associated with high arousal, such as group encounters. Although wild and captive sifakas exhibit the same alarm call system and use the same alarm call types, we discovered striking differences in the usage and perception of some of the alarm calls. We argue that these differences indicate either an evolutionary drift in the meaning of these calls or reflect cultural variation. The latter possibility is consistent with our understanding of the ontogeny of call usage and comprehension.