AnSTRACT.--Do songs of songbirds, which learn to sing, provide reliable clues to genetic identity in zones of secondary contact? How do some songbird species maintain such highly stereotyped songs throughout extensive geographic ranges? These two questions were addressed with a study of song development by Carolina and Black-capped chickadees (Parus carolinensis and P. atricapillus). In one hand-reared, mixed group in the laboratory, male Carolina Chickadees produced better imitations of a tape-tutored Black-capped Chickadee fee-bee song than did two male Black-capped Chickadees. In another mixed group, a male Black-capped Chickadee produced a better imitation of tape-tutored Carolina Chickadee song elements than did the Carolina Chickadee males themselves. Black-capped Chickadees in an additional experiment were tutored with normal fee-bee songs and with fee-fee, bee-bee, and bee-fee songs; these males also produced highly abnormal songs, although songs of males within groups converged on one another, reinforcing ideas that social interactions are crucial for the song learning process. These data thus reveal that song in secondary contact zones of these chickadees is probably not a good indication of genotype. The feat of Black-capped Chickadee song stereotypy in nature, together with other features of their singing behavior (e.g. social and hormonal determinants of singing, subsong by both sexes but loud songs only from males), remain both puzzling and fascinating. SONGS OF closely related bird species, especially in allopatry, are usually unmistakable to human listeners. Thus, throughout most of its geographic range, the Black-capped Chickadee (Parus atricapillus) utters a remarkably stereotyped fee-bee song (actually fee-bee-ee; see below), a two-noted whistle, with the first whistle slightly higher in frequency than the second (review in Hailman 1989). Likewise, the song of the congeneric Carolina Chickadee (P. carolinensis) is immediately recognizable. It typically consists of an even number of whistles, often four, with each odd-numbered whistle higher in frequency than the immediately following even-numbered whistle (Ward and Ward 1974).
Songs of each male and songs from location to location seem more variable than do the songs of theBlack-capped Chickadee, yet nowhere in allopatry would one confuse songs of the two species. Such stereotypy often deteriorates, however, in contact zones with a closely related congener (e.g. Helb et al. 1985). Among species that do not imitate their vocalizations from other individuals (e.g. suboscine flycatchers; Kroodsma and Konishi 1991), vocal displays can provide good evidence of genetic identity (Johnson and Marten 1988). Because the songs of chickadees and other songbirds are cultural imitations (e.g. Slater 1989), however, the existence of intermediate or hybrid songs does not indicate that birds are genetic hybrids. For some songbirds, plumage differences between sister taxa provide visible genetic evidence of hybridization. Among the buntings (Passerina spp.), gro...