In a population of chaffinches (Fringilla coelebs) some song types are produced by many individuals while others occur only in the repertoire of single individuals. This frequency distribution of song types fits that predicted from a model for the frequency of neutral alleles, suggesting that song types are copied at random rather than some being favoured over others. Both comparison with this model and computer simulations suggest that the mutation in song copying is around 15%, changes in this case arising either through immigration or because of inaccuracies in copying. Such a rate of change would also lead the songs present to change with time to much the same extent as we have found. No evidence could be found that particular combinations of song types tended to occur more often than expected in the same repertoire, suggesting that repertoires are built up by copying from more than one individual. Rare songs (those unique to a particular individual in the population) do, however, tend to cluster in the same repertoire, probably because some of the birds possessing them are immigrants. Repertoire size clusters at around 2-3 song types but is not significantly different from that predicted if birds can learn the same song type more than once so that their observed repertoire size is smaller than that which they actually possess.
No takes yet. Share an insight, caveat, or question.
Slater et al. (1980) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: