behavior is still rudimentary. A direct approach is to block the change in a
single parameter while holding all other parameters constant and observe
concomitant changes in a behavior. However, this approach is not available
for many neural characteristics, such as anatomical parameters (e.g., one
cannot selectively block the changes in the number of synapses without
changing the dendritic arborization). In these cases, comparison across
species as well as sexes and individuals will be a fruitful approach to identify which of the neural characters holds the key to the vocal activation
(Brenowitz 1997). Alternatively, one could take a bottom-up approach and
focus on neurons with an identified function, such as motor neurons. By
identifying molecular and cellular changes induced by activating hormones
in these lower-order neurons in the vocal pathway that directly control
muscles, rather than higher-order cells such as telencephalic neurons,
we can obtain a better handle on how particular changes translate to
behavior.
3.1.2. Frogs
Many species of frogs vocalize in the context of courtship. As in birds, males
typically emit advertisement calls to attract females as well as to defend
their territories during the breeding season. Advertisement calls of males
are largely dependent on T and/or its metabolites. Circulating levels of
T fluctuate seasonally in frogs in the temperate zone (Licht et al. 1983;
Pierantoni et al. 1984). In these species, calling rates correlate well with seasonal elevations of plasma androgen (Townsent and Moger 1987; Solis and
Penna 1997). Castrating males abolishes calling behavior, whereas administering T or DHT reinstates calling (Schmidt 1966; Palka and Gorbrnan
1973; Kelley and Pfaff 1976; Wetzel and Kelley 1983). Moreover, exogenous
testosterone induces male-like courtship vocalizations in castrated female
African clawed frogs, Xenopus laevis (Hannigan and Kelley 1986). These
results suggest that androgen is a major hormone activating courtship vocalizations in male frogs.
In contrast to these results, however, studies in some species have shown
limited potency of androgen in activating advertisement calling in males.
For example, circulating levels of T show poor correlation with calling
behavior in male bullfrogs, Rana catesbeiana (Mendonça et al. 1985), and
systemic administration of T fails to activate calling behavior in both
castrated male leopard frogs, Rana pipiens, and Mexican leaf frogs,
Pachymedusa dacnicolor (Schmidt 1966; Palka and Gorbman 1973; Wada
and Gorbrnan 1977a, 1977b; Rastogi et al. 1986). These observations led
Moore (1978) to hypothesize that testicular androgens are necessary, but
not sufficient, for the activation of some sexual behaviors in amphibians;
some nontesticular hormones, in addition to T, are required.
Arginine vasotocin (AVT), a nonmammalian homolog of arginine vasopressin (AVP), may be such a hormone. AVT is a neuropeptide synthesized
6. Hormonal Control of Communication
291
single parameter while holding all other parameters constant and observe
concomitant changes in a behavior. However, this approach is not available
for many neural characteristics, such as anatomical parameters (e.g., one
cannot selectively block the changes in the number of synapses without
changing the dendritic arborization). In these cases, comparison across
species as well as sexes and individuals will be a fruitful approach to identify which of the neural characters holds the key to the vocal activation
(Brenowitz 1997). Alternatively, one could take a bottom-up approach and
focus on neurons with an identified function, such as motor neurons. By
identifying molecular and cellular changes induced by activating hormones
in these lower-order neurons in the vocal pathway that directly control
muscles, rather than higher-order cells such as telencephalic neurons,
we can obtain a better handle on how particular changes translate to
behavior.
3.1.2. Frogs
Many species of frogs vocalize in the context of courtship. As in birds, males
typically emit advertisement calls to attract females as well as to defend
their territories during the breeding season. Advertisement calls of males
are largely dependent on T and/or its metabolites. Circulating levels of
T fluctuate seasonally in frogs in the temperate zone (Licht et al. 1983;
Pierantoni et al. 1984). In these species, calling rates correlate well with seasonal elevations of plasma androgen (Townsent and Moger 1987; Solis and
Penna 1997). Castrating males abolishes calling behavior, whereas administering T or DHT reinstates calling (Schmidt 1966; Palka and Gorbrnan
1973; Kelley and Pfaff 1976; Wetzel and Kelley 1983). Moreover, exogenous
testosterone induces male-like courtship vocalizations in castrated female
African clawed frogs, Xenopus laevis (Hannigan and Kelley 1986). These
results suggest that androgen is a major hormone activating courtship vocalizations in male frogs.
In contrast to these results, however, studies in some species have shown
limited potency of androgen in activating advertisement calling in males.
For example, circulating levels of T show poor correlation with calling
behavior in male bullfrogs, Rana catesbeiana (Mendonça et al. 1985), and
systemic administration of T fails to activate calling behavior in both
castrated male leopard frogs, Rana pipiens, and Mexican leaf frogs,
Pachymedusa dacnicolor (Schmidt 1966; Palka and Gorbman 1973; Wada
and Gorbrnan 1977a, 1977b; Rastogi et al. 1986). These observations led
Moore (1978) to hypothesize that testicular androgens are necessary, but
not sufficient, for the activation of some sexual behaviors in amphibians;
some nontesticular hormones, in addition to T, are required.
Arginine vasotocin (AVT), a nonmammalian homolog of arginine vasopressin (AVP), may be such a hormone. AVT is a neuropeptide synthesized
6. Hormonal Control of Communication
291