Song overproduction, selective attrition and song dialects in the white-crowned sparrow

Male Puget Sound white-crowned sparrows, Zonotrichia leucophrys pugetensis, form large vocal dialects along the Pacific northwest coast of North Ameri ca. Most adult males sing a single song dialect throughout their lives. To determine how dialects are maintained in this species, I studied two popula tions at the Columbia River mouth in a contact zone between two dialects. S ingers of each of the two dialects clustered together in space, forming loc al song 'neighbourhoods'. Two hypotheses for the maintenance of dialects we re tested. The late acquisition hypothesis predicts that yearling or adult male immigrants memorize their song from the territory neighbours they sett le next to. The selective attrition hypothesis predicts that males memorize a variety of dialects early in life, overproduce song dialects upon arriva l, and later selectively retain the one prelearned dialect that best matche s what their neighbours sing. In 1997 and 1998, 35-40% of new territory occ upants sang two dialects upon arrival in April, and then over the course of days to several weeks, discarded one dialect from their repertoire. Fourte en of 16 (88%) kept as their adult song the dialect that matched the dialec t sung by the majority of their neighbours. No male added a new dialect to his repertoire after arrival, nor did males alter their retained dialect to more closely resemble their neighbours' songs. New arrivals that overprodu ced dialects upon arrival were significantly more likely to match their nei ghbours' dialect than males that did not overproduce upon arrival. In a pla yback experiment, males in the overproduction stage engaged in matched coun tersinging to the dialect played to them. These observations and the experi ment support the selective attrition hypothesis: males visit and memorize a variety of dialects, probably in their hatching-year summer, overproduce d ialects upon arrival the next spring, and then selectively retain the one d ialect that matches the local song culture. Vocal plasticity late in life i s not the result of a late sensitive phase for song memorization, but rathe r results from behavioural selection operating on a pre-existing repertoire of song dialects. (C) 2000 The Association for the Study of Animal Behavio ur.