Social context influences androgenic effects on calling in the green treefrog (Hyla cinerea)

Courtship behavior in frogs is an ideal model for investigating the relatio nships among social experience, gonadal steroids, and behavior. Reception o f mating calls causes an increase in androgen levels in listening males, an d calling, in turn, depends on the presence of androgens. However, previous studies found that androgen replacement does not always restore calling to intact levels, and the relationship between androgens and calling may be c ontext dependent. We examined the influence of androgens on calling behavio r in the presence and the absence of social signals in male green treefrogs (Hyla cinerea). We categorized calling during an acoustic stimulus (mating chorus or tones) as evoked and calling in the absence of a stimulus as spo ntaneous. Intact males received a cholesterol implant, castrated males were castrated and received a cholesterol implant, and T-implanted males were c astrated and received a testosterone implant. The androgen levels (mean +/- SE ng/ml of plasma) achieved by the implants were as follows: castrated ma les, 1.2 +/- 0.2; intact males 21.9 +/- 7.0; T-Implanted males, 254.6 +/- 3 9.5. As in other frogs, calling depends on the presence of androgens, as ca stration abolished and T replacement maintained calling. However, among int act and T-implanted males, the influence of androgens on calling differed b etween spontaneous and evoked calling. There was a positive effect of andro gen treatment on spontaneous call rate and a positive correlation between s pontaneous call rate and androgen levels. The influence of androgen levels on evoked call rate was more complex and interacted with acoustic treatment . Surprisingly, T implants suppressed the chorus-specific increase in calli ng that is evident in intact males. In addition, in response to the chorus, T-implanted males called less than did intact males, in spite of higher an drogen levels. Furthermore, variation in androgens did not explain variatio n in evoked call rate. These data indicate that androgens influence the mot ivation to call, but that, when socially stimulated, androgens are necessar y but insufficient for calling. (C) 2001 Elsevier Science.