This is, to our knowledge, the first report of cuckoo hosts physically ejecting cuckoo nestlings from their nests, although nestling rejection (i.e. nest abandonment by hosts containing parasite young) has been reported previously (Grim et al. 2003
; Langmore et al. 2003
). We speculate that this ejection behaviour is an anti-parasitism strategy, for the following reasons. First, at least two confirmed nest owners ejected live nestlings from their nest, suggesting that this behaviour was not infanticide by intruders. Second, the morphological similarity between the nestling of the cuckoo and that of its host () suggests that this is a consequence of the host's ability to discriminate odd looking cuckoo nestlings, comparable to the outcome seen in cuckoos mimicking hosts' eggs to avoid detection (Brooke & Davies 1988
However, the present study did not provide sufficient evidence to show that the host adults rescued progeny from host-evicting cuckoo young. Rather, some of them killed their own young, but this is to be expected since similar recognition errors are known in egg-ejecting host species (e.g. Davies & Brooke 1988
). Therefore, we need to examine whether the hosts are able to achieve higher fitness by ejecting cuckoo nestlings even with the cost of ejecting their own young.
The evolutionary trajectory of this probable anti-parasitism strategy at the nestling stage may be quite different from that of abandonment of parasitized nests at the same stage. Langmore et al. (2003)
suggested that defence mechanisms at the nestling stage would evolve only after host defence at the egg stage had been breached by the parasite. Interestingly, our hosts seem to lack any anti-parasitism strategy at the egg stage (the cuckoo egg does not mimic host eggs, ). In fact, we observed only one case of egg ejection in our 4 years of research, and this happened when the egg did not hatch after the full incubation period (, no. 5).
These factors suggest that the host may have bypassed the egg rejection strategy and went straight to the evolution of nestling ejection strategy. The lack of egg rejection, and the evolution of nestling ejection, may be owing to a coincidence, or constraints such as small bill size (Rohwer & Spaw 1988
), physical structure of the nest, e.g. domed nest (Brooker et al. 1990
; Langmore et al. 2009b
) or some other flexible strategies by the host (Langmore et al. 2009a
). Future research on this apparently unique system may give us new insights into the coevolution of avian brood parasitism.