In songbirds, the size of brain nuclei that control song learning and production change seasonally. These changes are mainly controlled by seasonal changes in plasma testosterone (T) concentration. One hypothesis to explain why it may be adaptive for these areas to regress in the fall is that this would decrease the metabolic demand of maintaining a large song system when singing is reduced or absent. We used a marker for cellular metabolism to examine birds with regressed song nuclei and compared them to birds whose song nuclei were induced to grow by administration of exogenous T. Photorefractory male Gambel's white-crowned sparrows were captured during their autumnal migration and kept in outdoor aviaries on a natural photoperiod. We implanted birds with Silastic capsules containing T or with empty implants. Three weeks later the birds were sacrificed. We assayed the brains for cytochrome oxidase (CO) activity and measured the volume of four song nuclei: HVc, RA, 1MAN, and area X. All four nuclei increased in volume in response to T treatment. T treatment increased the metabolic capacity of area X, HVc, and RA relative to surrounding tissue but had no effect on the metabolic capacity of 1MAN. These results support the hypothesis that song nuclei are more metabolically active under the influence of T than they are when plasma T levels are low.

In seasonally breeding songbirds, brain nuclei of the song control system that act in song perception change in size between seasons. It has been hypothesized that seasonal regression of song nuclei may impair song discrimination. We tested this hypothesis in song sparrows (Melospiza melodia), a species in which males share song types with neighbors and must discriminate between similar songs in territorial interactions. We predicted that song sparrows with regressed song systems would have greater difficulty in discriminating between similar songs. Sparrows were housed either on short days (SD) and had regressed song circuits, or were exposed to long days and implanted with testosterone (LD+T) to induce full growth of the song circuits. We conducted two experiments using a GO/NO-GO operant conditioning paradigm to measure song discrimination ability of each group. Birds learned four (experiment 1) or three (experiment 2) pairs of song types sequentially, with each pair more similar in the number of shared song elements and thus more difficult to discriminate. Circulating T levels differed between the SD and LD+T groups. The telencephalic song nuclei HVc, RA, and area X were larger in the LD+T birds. The two groups of sparrows did not differ, however, in their ability to learn to discriminate between shared song types, regardless of the songs' similarity. These results suggest that seasonal changes in the song control system do not affect birds' ability to make difficult song discriminations.