To what extent has the style of song development among songbirds coevolved
with other life history strategies? Among Cistothorus wrens in North Americ
a, it seems that sedentary or site-faithful habits of marsh wrens, C. palus
tris, favour song imitation, but seminomadic habits of sedge wrens, C. plat
ensis, favour song improvisation, whereby each male generates a large but u
nique song repertoire. In this study, we tested whether more sedentary popu
lations of sedge wrens in the Neotropics would imitate songs. At our primar
y study site near Cartago, Costa Rica, breeding birds were colour-banded du
ring 1995 and 1996, and follow-up surveys revealed that the birds remained
at this site the year round. Extensive tape recording and analysis of songs
showed that males had large song repertoires (200-390+.songs), and that ma
ny songs were shared among neighbouring males. In addition, males only 27 k
m distant, at La Pastora, used different songs. Furthermore, matched counte
rsinging, in which two males answer each other with identical song types, w
as recorded near Brasilia, in Brazil. The sharing of songs among permanent
neighbours, microgeographical variation in song, and matched countersinging
can be achieved only through song imitation, thus revealing a striking dif
ference in the style of song development among different populations of the
sedge wren. Tn the Neotropics, having predictable neighbours throughout li
fe appears to have favoured song imitation, so that individuals can interac
t using a common, learned code typical of the local population; among more
mobile populations in North America, however, individuals improvise large r
epertoires of species-typical songs, thereby enabling singing males to comm
unicate with any individual, no matter what the population of origin. Strat
egies of song development must correlate with life history features, and fu
rther surveys are needed to make sense of the great diversity of singing be
haviours among songbirds. (C) 1999 The Association for the Study of Animal
Behaviour.