Acoustic communication in two groups of closely related treefrogs

Studies of vocal behavior in anurans provide significant insights about the mechanisms and evolution of animal communication. Females extract the information required for a mating decision from relatively stereotyped advertisement calls that are produced by males. Aggressive signals, used in exchanges between males, are more variable in form, as might be expected of communication between rivals, but overall, their structure is more similar in closely related species than is that of advertisement calls. One open question is why, within the groups of treefrogs considered in this review, aggressive signaling is so prevalent in some species and rare or absent in close relatives.

Comparative studies of the acoustic criteria used in signal selection reveal significant differences among species. What are important properties for one species are seemingly irrelevant for its closest relative. Moreover, as exemplified by gray treefrogs and European katydids, even when females apparently base preferences on the same acoustic property, deeper analysis reveals that different mechanisms are used by each species to extract a different aspect of that property. In several of the treefrog species considered in this review, some preferences and preference-mechanisms are more divergent than call structure (Sections II.E and III.E). More research with additional species is needed before we can even attempt to make broad generalizations about the coevolution of signalers and receivers in animal communication systems.

Communication usually takes place in acoustically difficult situations: dense choruses, where background noise presents special challenges for signal detection and pattern recognition. Although they can quantify acoustic preferences, playback experiments in simplified laboratory situations are unlikely to accurately predict patterns of mate choice in nature. Studies with species in the green and gray treefrog groups show that the optimum values of particular acoustic properties can vary depending on the absolute playback level to which alternatives are equalized, the relative playback levels of alternatives, the number of sounds presented and their timing relationships, and the ambient temperature. Comparable results are available for other anurans and acoustic insects. Technology that allows researchers to monitor groups of calling males, whose signals and mating success can be quantified, promises to refine our estimates of how female selectivity contributes to mate choice in nature (e.g., Passmore et al., 1992; 156 and 157).

Studies of geographical variation in communication systems are limited because we know so little about patterns of variation in receiver selectivity. Even where acoustic properties have diverged in sympatric areas relative to allopatric areas, and hence show the pattern expected of reproductive character displacement, we often do not always know whether these properties are even relevant to females. A few studies have shown that female selectivity diverges in sympatric areas, whereas signal structure does not, a result that may reflect the fact that the choosing sex usually has more to lose from mating mistakes than the signaling sex. I predict, however, that studies of geographical variation in the acoustic criteria and selectivity of receivers will ultimately provide significant insights concerning the evolution of communication systems that will go well beyond the narrow issue of reproductive character displacement.