Introduction

Animals communicate by emitting and exchanging a wide variety of signals. Many of these signals are unisensory, in that they are received by one sensory system in the perceiving animal. These unisensory signals can be highly effective, as the glorious visual display of the male peacock's tail feathers illustrates. However, in many cases animals do more than just emit a signal in one channel: they simultaneously emit signals in multiple channels. For example, male red jungle fowl display by spreading their colorful wing feathers and simultaneously running a foot through the feathers, creating a staccato auditory signal that accompanies the visual display (Fig. 14.1; Kruijt, 1962). This type of display can be called a multisensory signal, because information travels through multiple sensory channels. Since sensory systems are characteristics of perceivers, we can more formally define multisensory signals as signals that can be received simultaneously via more than one sensory system of a perceiver.

Figure 14.1.  

Courtship display of male Burmese red jungle fowl (after Kruijt, 1962). (A) Male circles around female. (B) Male stretches out his wing and runs his raised foot through the primary feathers of the wing, creating a staccato sound that accompanies the visual display. (Adapted with permission from Kruijt, J. P. [1962]. On the evolutionary derivation of wing display in Burmese red jungle fowl and other gallinaceous birds. Symposia of the Zoological Society of London, 8, 25–35. © 1962 The Zoological Society of London.)

Multisensory signals are very common in natural animal behavior. Like unisensory signals, multisensory signals are used to convey various messages in a variety of situations. An animal's choice of signal channel depends on both the biotic and the abiotic environment in which it lives. Mole rats, for example, live underground, so in addition to tactile and olfactory signals, they make copious use of vibratory signals that capitalize on their subterranean environment. By contrast, most birds use visual and acoustic signals that travel well through air. Marler (1959) and Endler (1992) list qualities of each sensory channel independently that make them well-suited for use in particular environments, and Rosenthal and Ryan (2000) have compared the visual and the acoustic modalities in particular. In some cases one sensory system is used in combination with another more frequently than the reverse. In penguins, for example, vocalizations are always associated with particular postures, but not all postures are associated with vocalizations (Jouventin, 1982).

The immediate biotic environment influences the choice of channel in a number of ways. A relatively quiet channel with the fewest competitors is best for effective communication. For example, many sympatric species share “acoustic space” by calling either at a different frequency (pitch) or at a different time of day than their neighbors. Recent evidence suggests that chimpanzees (Pan troglodytes) may be able to choose which channel to use (vocal or visual) depending on the attentional state of their audience (i.e., whether or not the audience can see them; Hostetter, Cantero, & Hopkins, 2001). Other influences of receivers on signaler behavior have been discussed by Guilford and Dawkins (1991) and by Rowe (1999).

If unisensory communication signals are effective, why add sensory channels and emit multisensory signals? In theory, if components are redundant (having the same meaning), the message has a higher chance of successful transmission despite a noisy communication channel; this benefit is often referred to as insurance (Rand & Williams, 1970; Wilson, 1975) or error reduction (Wiley, 1983). If components are nonredundant (having different meanings), the display may carry additional information per unit time (e.g., Johnstone, 1995), or its meaning may be modified by the co-occurrence of the multiple channels (Hughes, 1996; Marler, 1967). In practice, what do multiple channels add to the communication event? In this chapter I discuss how multisensory signals enhance animal communication with examples from animals in a wide variety of taxa. Although I do not directly address multicomponent single-channel signals here, much of the following discussion is relevant to the literature on multiple visual signals (e.g., Zucker, 1994; Zuk, Ligon, & Thornhill, 1992) and on complex pheromone blends (e.g., Borden, Chong, & Lindgren, 1990; Johnston, Chiang, & Tung, 1994).

I first provide a brief overview of descriptive observational work documenting multisensory communication in animals. Because the literature on multisensory communication is highly varied in respect to species, methods, and outcomes, I present a framework for classifying multisensory signals in order to help organize types of signals and outcomes that occur in nature (see Partan & Marler, 1999). Most of this chapter is devoted to developing this classification system in detail. The classification is based on the relationship among the signal components. To assess this relationship, I rely on experimental studies that examine the behavioral consequences of signaling in one channel versus signaling in multiple channels and consider what, if any, additional information is provided by the additional channels. To understand how multisensory signals are used in communication we must measure the responses of animals to both the unisensory and multisensory conditions.

Multisensory processes are important for behavior at many levels and in myriad realms. The chapter concludes with a brief review of many other ways, in addition to communication, that multisensory processes play a role in animal lives.