Extraordinary variation in a simple song: No geographical patterns in initial phrase variation of the Yellowhammer, a passerine with pronounced dialects

Geographical variation of birdsong is used to study various topics from
cultural evolution to mechanisms responsible for reproductive
barriers or song acquisition. In species with pronounced dialects,
however, patterns of variation in nondialect parts of the song
are usually overlooked. We focused on the individually variable initial
phrase of the song of the Yellowhammer (Emberiza citrinella), a
common Palearctic passerine which became a model species for dialect
research. We used a quantitative method to compare the similarity
of initial phrases from the repertoires of 237 males recorded at
different spatial scales in a central European country covering all main
dialect types. We hypothesized that patterns of initial phrase
sharing and/or phrase similarity are affected by dialect boundaries and
geographical proximity (i.e. that birds from the same dialect
regions use more similar phrases or share them more often). Contrary to
our expectations, initial phrase variation seems unrelated to
dialects, as we did not find higher similarity either among recordings
from the same dialect areas or among those from the same
locality. Interestingly, despite the immense variability of
phrase types detected (only 16% of 368 detected initial phrase
types were shared between at least 2 males), a relatively
high proportion of males (45%) was involved in sharing, and males
using the same initial phrase were located anywhere from tens of
meters to hundreds of kilometers apart. The patterns of variation suggest
that precise copying during song learning as well as
improvisation play important roles in forming individual repertoires in
the Yellowhammer. Our data also confirm previous indications that
the repertoires of Yellowhammer males (i.e. the composition of initial
phrases) are individually unique and temporally stable. This
makes the species a good candidate for individual acoustic
monitoring, useful for detailed population or behavioral studies
without the need for physical capture and marking of males.