9701351 Beecher Bird song is widely recognized as a model system for investigations into the biological basis of learning. It provides the best example to date of natural learning in a vertebrate, with many striking parallels to human language learning. Moreover, in recent years many discoveries concerning the brain mechanisms underlying song learning have suggested that this model system may reveal much about the general processes of learning in the brain. This proposal will examine one key aspect of song learning, the tendency of young birds to memorize more songs than they will ultimately keep for their final repertoire of songs. Most studies have been done in the laboratory, but the major study will be conducted on a free-living population of song sparrows, thus permitting evaluation of the adaptive significance and ecological context of song learning. The major question addressed is why do male song sparrows produce more song types during juvenile song than they have in their final adult repertoires? Specifically, why do they keep certain songs and not others. The hypothesis is that they keep those songs that best match those of their neighbors during their first breeding season. The research will record "plastic" songs of juvenile males (song is not fully formed until March or April of the bird's first spring) and compare these to "crystallized" song types in the bird's adult repertoire and to the songs of adult males present in the bird's natal summer and his first spring (a partially overlapping group). A parallel study in the laboratory of hand-raised birds tutored by several adult males will also be conducted. This research will allow a better understanding of how song is learned, and insights into learning processes in general.
