/div>
Abstract
Songbirds learn to sing by imitating their conspecific songs through social learning. It is commonly thought that in species with small repertoires, so‐called crystallization of the song repertoire takes place before the first breeding attempt and afterwards their repertoires remain unchanged. However, the number of studies in which individual song repertoires have been tracked longitudinally under natural conditions is still small. The ortolan bunting (Emberiza hortulana) is a small Eurasian passerine species. Studies on this species have shown that males have small song repertoires (usually 2–3), share the same final phrase within a local dialect area, and their repertoires remain unchanged during their lifetimes. We studied the whole, isolated Norwegian population of this species with 100–150 individually marked males. We compared repertoire sizes and contents using cross‐sectional and longitudinal approaches. We focused on marked males aged 2–10 years and recorded in up to five breeding seasons between 2001 and 2006. There was no local dialect because songs with different final phrases occurred both within and between males. Repertoire size varied between 1 and 24 (4.7 ± 3.4) song types per male. Longitudinal analysis showed an increase in repertoire size in 29% and a decrease in 21% of males. Males whose repertoires remained the same in size between years often substantially changed in contents. New song types appeared in almost 70% of males, and over 30% of males started singing new syllable types after the second calendar year. In almost 50% of males, new song types appeared that were built as new combinations of syllables shown in previous years. In 30% of males, the song complexity increased with age. Simultaneously, over 60% of males selectively ceased singing some song types. We suggest that these repertoire alterations are linked to the isolation and fragmentation of this population, which may affect song learning patterns through increased breeding dispersal and a strongly male‐biased sex ratio. Similar studies from abundant and continuous populations of the species are needed to test this idea.
