the most dramatic sex differences in neuronal structures are due to the
developmental differentiation process, sexually distinct levels of plasma
androgen in adulthood also contribute to the differences in the song-nuclei
volume. Implanting T into adult female canaries increases the size of HVc
and RA (Nottebohm 1980; DeVoogd et al. 1985), and the same treatment
in adult male zebra finches increases the volume of nXIIts and DLM
(Gurney and Konishi 1980; Arnold 1980b; Gurney 1982). Thus, circulating
androgen can bind to receptors in the neurons of some song-control nuclei
and subsequently increase their volume in adults, although exactly how
these changes occur, in terms of gene expression, is not yet clear.
How are androgen-dependent increases in the volume of song-control
nuclei achieved at the cellular level? Fluctuations in the volume of vocal
nuclei derive from changes in the number, density, and size of neurons. In
a few species examined, the number and density of neurons in song-control
nuclei increase during the spring and decrease in the fall, and the somal size
and dendritic arborization of each neuron also increase during the spring
in adult males, suggesting that these parameters are controlled by circulating androgen (DeVoogd et al. 1985; Clower et al. 1989; Brenowitz et al. 1991;
Hill and DeVoogd 1991; Smith et al. 1995). In canaries, the cell bodies of
male RA neurons are larger and their dendrites are more numerous and
cover a larger area than those of female neurons, whereas the number of
neurons does not differ between the two sexes (DeVoogd and Nottebohm
1981a). Treating gonadectomized female canaries with T results in growth
of the dendritic field of RA neurons (DeVoogd and Nottebohm 1981b).
Taken together, androgen-dependent increases in the volume of songcontrol nuclei result from the increases in the number, density, and/or the
size of neurons within each nucleus.
What subcellular changes are associated with androgen binding in
neurons of song-control nuclei? One clear example occurs at the synaptic
level. DeVoogd (DeVoogd et al. 1985) found that injecting T into adult
female canaries increases the number of synapses within RA. The synapses
of T-treated females had shapes characteristic of stronger synapses, and the
number of synaptic vesicles found in the synaptic terminal was dramatically
increased compared with control females. This result strongly suggests that
new and stronger synapses are formed in response to T in this vocal nucleus.
So far, we have seen that androgen activates courtship singing in male
songbirds, binds to neurons in song-control nuclei, increases the volume of
the nuclei by increasing the number, size, and density of neurons, and modifies the strength of synapses within song nuclei. These pieces of information make up an extended catalog of bivariate observations with hormone
as an independent variable and song production or neuronal structure as
dependent variables. However, it is difficult to identify which of the structural changes in the CNS are actually involved in vocal production, and how
these changes mediate expression of behavior (Fig. 6.6). What we have to
keep in mind is that not all physiological and structural changes induced by
activating hormones in the CNS or the vocal organ can be directly related
6. Hormonal Control of Communication
289
developmental differentiation process, sexually distinct levels of plasma
androgen in adulthood also contribute to the differences in the song-nuclei
volume. Implanting T into adult female canaries increases the size of HVc
and RA (Nottebohm 1980; DeVoogd et al. 1985), and the same treatment
in adult male zebra finches increases the volume of nXIIts and DLM
(Gurney and Konishi 1980; Arnold 1980b; Gurney 1982). Thus, circulating
androgen can bind to receptors in the neurons of some song-control nuclei
and subsequently increase their volume in adults, although exactly how
these changes occur, in terms of gene expression, is not yet clear.
How are androgen-dependent increases in the volume of song-control
nuclei achieved at the cellular level? Fluctuations in the volume of vocal
nuclei derive from changes in the number, density, and size of neurons. In
a few species examined, the number and density of neurons in song-control
nuclei increase during the spring and decrease in the fall, and the somal size
and dendritic arborization of each neuron also increase during the spring
in adult males, suggesting that these parameters are controlled by circulating androgen (DeVoogd et al. 1985; Clower et al. 1989; Brenowitz et al. 1991;
Hill and DeVoogd 1991; Smith et al. 1995). In canaries, the cell bodies of
male RA neurons are larger and their dendrites are more numerous and
cover a larger area than those of female neurons, whereas the number of
neurons does not differ between the two sexes (DeVoogd and Nottebohm
1981a). Treating gonadectomized female canaries with T results in growth
of the dendritic field of RA neurons (DeVoogd and Nottebohm 1981b).
Taken together, androgen-dependent increases in the volume of songcontrol nuclei result from the increases in the number, density, and/or the
size of neurons within each nucleus.
What subcellular changes are associated with androgen binding in
neurons of song-control nuclei? One clear example occurs at the synaptic
level. DeVoogd (DeVoogd et al. 1985) found that injecting T into adult
female canaries increases the number of synapses within RA. The synapses
of T-treated females had shapes characteristic of stronger synapses, and the
number of synaptic vesicles found in the synaptic terminal was dramatically
increased compared with control females. This result strongly suggests that
new and stronger synapses are formed in response to T in this vocal nucleus.
So far, we have seen that androgen activates courtship singing in male
songbirds, binds to neurons in song-control nuclei, increases the volume of
the nuclei by increasing the number, size, and density of neurons, and modifies the strength of synapses within song nuclei. These pieces of information make up an extended catalog of bivariate observations with hormone
as an independent variable and song production or neuronal structure as
dependent variables. However, it is difficult to identify which of the structural changes in the CNS are actually involved in vocal production, and how
these changes mediate expression of behavior (Fig. 6.6). What we have to
keep in mind is that not all physiological and structural changes induced by
activating hormones in the CNS or the vocal organ can be directly related
6. Hormonal Control of Communication
289