The Evolution of Animal Communication: Reliability and Deception in Signaling Systems

5 Honesty and Deception in Communication Networks Our analysis of honesty and deception thus far has taken as its starting point an implied view of communication as a fundamentally dyadic interaction, with a sender and a receiver that may differ in their evolutionary interests, but which nonetheless interact with each other independently of the influences of other actors. This dyadic view of communication has provided an appropriate platform for our discussion, for two reasons. First, viewing communication as a dyadic interaction has long been the dominant perspective in studies of animal communication (Marler and Hamilton 1966, Brown 1975, Wilson 1975), and most of the literature has taken this perspective as its starting point. Second, the dichotomy between sender and receiver defines the evolutionary conflict of interest in a signaling interaction and thus captures the essential issue underlying reliability and deceit. Nevertheless, it is commonly—perhaps almost invariably —the case that a signal will be detected by more than one receiver. McGregor and his colleagues have introduced the idea of “communication networks” to describe the broader social environment in which signaling may occur (McGregor 1993, McGregor and Dabelsteen 1996, McGregor and Peake 2000, McGregor et al. 2000). The costs and benefits of a signaling interaction may not be altered substantially by adding extra receivers to the mix, if all of those receivers fall into the same functional class. When several female crickets detect the mate-attraction call of a single male, or when both parent birds detect the begging calls of chicks in a nest, the analysis of costs and benefits may be more complicated, because there are additional individuals to consider, but the fundamental equation does not necessarily change. In other cases, however, the different potential receivers of a signal may introduce different selective pressures. Signals produced in an interaction between two territorial males, for example, are likely to be detected by other males not involved in the interaction, as well as by females (McGregor 1993, McGregor et al. 2000). The exchange between the two signaling males may help to determine the outcome of the aggressive interaction between them, and the honesty of signaling will be enforced by a receiver-dependent cost (see chapter 4). But what of the other potential receivers ? Both females and other males not directly involved in the dyadic interaction may gain information about the two interacting individuals, and change 182 â CHAPTER 5 their future behavioral responses to one or both males on the basis of that information. The potential costs and benefits to both the signaler and the receiver in the dyadic interaction must include the effects of the signaling interaction on these other, “third-party” receivers in the network. The additional costs and benefits so produced may affect the mix of reliability and deceit found in the signaling system at evolutionary equilibrium. Third-Party Receivers The idea that individuals of other species can listen in on and exploit signaling exchanges among conspecifics is long established (Marler 1955, Otte 1974, Myrberg 1981). Indeed, much early debate in the animal-communication literature centered on the question of whether a signal intercepted by an unintended receiver of a different species constitutes “true” communication (e.g., Marler 1977, reviewed in Bradbury and Vehrencamp 1998). Is it communication, for example, when a gecko finds its prey by orienting to the calling of male crickets (Sakaluk and Belwood 1984)? Or when a parasitoid fly locates an ant host by following its alarm pheromones (Feener et al. 1996)? Whether or not the exploitation of signals by natural enemies is considered communication, thirdparty receivers such as these clearly may impose serious costs on both the signaler and the intended receiver in a signaling interaction. We have previously considered detection by predators as a potential cost for signals such as the begging of nestling birds and the carotenoid coloration of guppies. Another, particularly well-described example is provided by the mateattraction call of the túngara frog (Physalaemus pustulosus). Male túngara frogs produce a complex call beginning with a frequency-modulated “whine” that may be followed by one or more harmonically rich “chucks.” The number of chucks produced by a male increases with his body size, and females respond more to calls having a greater number of chucks, suggesting that the number of chucks is an honest indicator of size used by females in mate choice (Rand and Ryan 1981, Ryan 1983). The honesty of this signaling system would be ensured if the chucks...