Passerine hosts of parasitic cuckoos usually vary in their ability to discr
iminate and reject cuckoo eggs. Costs of discrimination and rejection error
s have been invoked to explain the maintenance of this within-population va
riability. Recently, enforcement of acceptance by parasites has been identi
fied as a rejection cost in the magpie (Pica pica) and its brood parasite,
the great spotted cuckoo (Clamator glandarius). Previous experimental work
has shown that rejecter magpies suffer from increased nest predation by the
great spotted cuckoo. Cuckoo predatory behavior is supposed to confer a se
lective advantage to the parasite because magpies experiencing a reproducti
ve failure may provide a second opportunity for the cuckoo to parasitize a
replacement clutch. This hypothesis implicitly assumes that magpies modulat
e their propensity to reject parasite eggs as a function of previous experi
ence. We tested this hypothesis in a magpie population Breeding in study pl
ots varying in parasitism rate. Magpie pairs that were experimentally paras
itized and had their nests depredated, after their rejection behavior had b
een assessed, changed their behavior from rejection to acceptance. The chan
ge in host behavior was prominent in study plots with high levels of parasi
tism, but not in plots with rare or no cuckoo parasitism. We discuss three
possible explanations for these differences, concluding that in study plots
with a high density of cuckoos, the probability for a rejecter magpie nest
of being revisited and depredated by a cuckoo is high, particularly for re
placement clutches, and, therefore, the cost for magpies of rejecting a cuc
koo egg in a replacement clutch is increased. Moreover, in areas with high
levels of host defense (low parasitism rate), the probability of parasitism
and predation of rejecter-magpie nests by the cuckoo is reduced in both fi
rst and replacement clutches. Therefore, rejecter magpies in such areas sho
uld not change their rejection behavior in replacement clutches.