


This is because most young birds learn the species-specific songs during the first year of their lives. With some exceptions (e.g., gray catbirds, sedge warblers, brown-headed cowbirds ), birds raised in acoustic isolation during the sensory phase produce atypical songs as adults. Territorial conflicts are energetically costly, and the song of a local population, or dialect, is thus thought to be favored as it can serve as an indication of regional origin, minimizing physically demanding defense measures. When maintaining a breeding territory, familiar neighbors are preferred over newcomers because newcomers, who do not yet have a territory, are more likely to expand their territories than established neighbors. On the other hand, the "song-sharing hypothesis" suggests that male-male competition selects for simpler, more homogenous songs (Beecher & Brenowitz 2005). In fact, in some species females prefer males with large repertoires, and males with larger repertoires have a higher reproductive success (i.e., they produce more offspring) than those with a smaller repertoire (Catchpole & Slater 2008). So what favors large and small repertoire size? According to the "repertoire hypothesis," a large repertoire is thought to be under directional sexual selection (Catchpole & Slater 2008, Beecher & Brenowitz 2005) where the choosy sex (often the females) favors one extreme end of a heritable, sexual trait. As we already learned, variation in repertoire size among species is extensive, and the characteristic of having a repertoire with multiple song types likely evolved more than once in songbirds (MacDougall-Shackleton 1997).
