Antm. Behav., 1984, 32, 280287 SONG LEARNING AND DECEPTIVE MIMICRY By P. K. McGREGOR & J. R. KREBS Edward Grey Institute, South Parks Road, Oxford, OX1 3PS, England Abstract. Payne (!978, 198ta, 1982) has suggested that young males of some bird species copy the songs of the previous occupant or the songs of local established males when setting up a territory and thereby increase their likelihood of success. According to this view, song copying is a form of decep- tive mimicry. We show that first-year great tits do not share more songs than expected with the pre- vious occupants of their territories, nor are shared songs used more often than other songs in their repertoires. Songs are shared with neighbours but not selectively with established neighbours and there is no difference in reproductive success between great tits sharing with established neighbours and with first-year neighbours. We discuss the general applicability of the deceptive mimicry hypothesis and suggest that song Sharing between neighbours may be advantageous for reasons other than deceptive mimicry. In this paper we present evidence from a long- term study of great tits (Parus major) relating to some recently-proposed hypotheses about the adaptive value of song learning (Payne 1978, 1981a, 1982, 1983) and song repertoires (Slater 1981), which are becoming widely cited, e.g. Rohwer (1982), Rich (1981) and Yasukawa et al. (1982), Payne has proposed two versions of an hypo- thesis for the adaptive value of song learning: (1) by copying the song of the previous owner of a territory, a young bird taking over that territory may be able to deceive others into re- acting as though the previous owner is still present, i.e. not attempting to displace the new- comer (Payne 1981a)', (2) in species with local dialects a young male may similarly give the deceptive impression of being an established male by copying the songs of older neighbours and thus conforming to the dialect group (Payne 1982). Again the consequence of the deception is that the newly established male is not attacked as often as it would be without the deception. According to both versions, flexible learning ability is advantageous in allowing a young male to adopt an appropriate song wherever it settles. Slater's (1981) hypothesis is similar to Payne's first proposal, but it refers to the frequency of use of songs in a repertoire rather than learning new songs. His idea is that a replacement bird taking over a terr!tory learns that the songs in its repertoire which resemble those of the previous owner are more effective at keeping out neighbours, therefore these songs come to be used most frequently. The evidence for Payne's first version of the hypothesis comes from his work on indigobirds, Vidua chalybeata (1981a). In this polygynous species there are local dialect groups of up to 20 males sharing song repertoires. Within each group there is one male that obtains most of the matings and when it disappears another bird takes over its call site. Payne proposes that the newcomer is more successful in take-overs if it copies the previous incumbent's songs. Evi- dence for copying is twofold: first, when new birds join a dialect group they adopt the local song pattern; second, when the most successful male adopts a new song type, the others copy it. This does not happen when another mate develops a new song. The second version of the hypothesis arises from Payne's work on indigo buntings (Passerina cyanea) in which small groups, usually of two to five birds share a song type. Payne (1982) found that males in a song neighbourhood, especially first-year males, who shared their song with older neighbours were more likely to mate and raise young than were non-sharers. Young indigo buntings learn their songs' from neigh- bours with which they interact at the start of their first season, and they sometimes subse, quently Change songs if they move to a new dialect area (Payne 198 lb). Since indigo buntings are capable of individual recognition (Emlen 1972) it is unlikely that young birds are mimick- ing the previous owner by copying the local dialect from neighbours. Instead Payne proposes that learning the local song type gives a new male art advantage by deceiving other young males into treating it as art established member of the neighbourhood. This is not the only possible interpretation of the data. Payne (1982) shows that first-year birds that share songs with adults tend to arrive at the breeding grounds earlier and also that the reproductive 280