Sexual difference in signal-receiver coevolution

The responses of males and females to a male signal were compared between two sister species, one in which the signal is present and one in which the signal is absent. The signal is a pattern of vertical bars found on males throughout the northern swordtail fish, genus Xiphophorus. In X. multilineatus, the bars function both to attract females and deter rival males. Males in X. nigrensis, the sister species to X. multilineatus, do not have bars. Xiphophorus nigrensis males did not respond differently to males with bars compared with males without bars, but X. nigrensis females were more attracted to males with bars than males without bars. Thus, the male response to bars is congruent in both taxa: male response was present when the male trait was present (in X. multilineatus) and absent when the male trait was absent (in X. nigrensis). The female response to bars, however, was congruent in X. multilineatus but incongruent in X. nigrensis: female response was present but the male trait was absent in X. nigrensis. This pattern suggests that of the three components of the communication system, (male response, female response and male signal), the signal and male response coevolved more closely than did the signal and female response. ? 1996 The Association for the Study of Animal Behaviour The evolution of extravagant male traits can be viewed from a conventional animal communication standpoint, in which male traits are signals and the males and females that respond to those traits are receivers (Boake 1991; Zahavi 1991; Zuk 1991; Harper 1992). It has been assumed that the evolution of a signal requires some type of coordination between signal and receiver; otherwise, changes in the signal would not be perpetuated (Alexander 1962; Butlin & Ritchie 1989). A growing number of studies, however, report cases in which female preference for a specific male trait is not congruent with the male trait. For example, female preference for male traits have been detected in taxa where males do not possess the trait (Ryan & Wagner 1987; Basolo 1990, 1995; Ryan & Rand 1993; Hill 1994). Patterns of signalreceiver coevolution have been of interest because they suggest the degree to which communication systems evolve as integrated and coordinated units, as in the case where the genetic bases for the signal and response are pleiotropic (Alexander 1962), as opposed to evolving more independently, as in the case of sensory exploitation where pre-existing preferences provide the selection that favours the evolution of certain signals (Ryan 1990, 1994; Ryan et al. 1990a). Male traits can evolve under sexual selection as signals used not only in attracting mates but also in deterring rivals. The idea that some male signals serve a dual function is not new (Tinbergen 1953) and has been demonstrated in numerous cases, such as visual signals in fish (Rowland 1989) and birds (von Schantz et al. 1989) and acoustic signals in frogs (Gerhardt 1994), birds (Catchpole & Slater 1995) and mammals (McComb 1987). In most sexual selection studies, however, the interaction between the male trait and the female preference, a dyad, is often investigated without reference to the male response to that signal (Boake 1991; Andersson 1994). If a signal influences both female and male responses, then studies of the function and evolution of the communication system should encompass the entire communication triad: the signal, the female’s response and the male’s response. By comparing all three components between taxa, the degree of coevolution of female response and male trait can be compared with the degree of coevolution of Correspondence M. R. Morris, Department of Biology, Montgomery College, Rockville, MD 20850, U.S.A. (email: [email protected]). M. J. Ryan is at the Department of Zoology, University of Texas, Austin, TX 78712, U.S.A. 0003–3472/96/111017+08 $25.00/0 ? 1996 The Association for the Study of Animal Behaviour