We tested the sexual responsiveness of female canaries, Serinus canari
a, to two sets of different types of female conspecific songs versus a
n adult male conspecific song. Female songs were either spontaneously
emitted ('S-songs') or were testosterone-induced ('T-songs'). Copulati
on-solicitation displays (CSD) were used as an index of female sexual
response. Playbacks were performed several days before and during egg
laying, a period of natural sexual responsiveness of the females to so
ng. We demonstrated that the weaker sexual displays of female canaries
were recorded to S-songs, thus suggesting that these types of female
songs do not contain fully functional sexual releasers. Three T-songs
elicited high levels of sexual displays, thus demonstrating that testo
sterone treatment map induce sexual release quality in the female song
s. Study of the phonology of these three T-songs strongly suggested th
at special song phrases map be good candidates as powerful sexual rele
asers. To test the sexual value of these female song phrases, we carri
ed out a third experiment, using hybrid songs where each of these spec
ial T-song phrase types was included in a well-known heterospecific co
ntext. Two phrases elicited high levels of sexual responses in females
. Essential features of the male full song, such as broadband rapid fr
equency modulations and high repetition rate, are retrieved in both fe
male song phrases. Taken together, these data demonstrate that testost
erone treatment not only induces a male-like structure in the songs of
females, but also induces functionally 'male-like' songs. This result
allows features of the vocal control network of testosterone-treated
females to be compared with those of adult males singing full songs, t
o distinguish neural correlates of testosterone-dependent full songs.
However, because testosterone does not induce functionally male-like s
ongs in all the females, neuroanatomical structure-function correlatio
ns need detailed behavioural analysis.