Predators may affect prey demography through direct predation and the costs of anti-predator behavioural responses. Though theory has long suggested that predator-induced changes in prey behaviour may reduce prey reproduction and increase deaths from other causes, there are almost no experimental studies on wild birds or mammals that have examined all of the steps in the pathway from predation risk to behavioural responses to changes in demography. We tested the effects of predation risk on parental anti-predator behaviours and demography in free-living song sparrows. Predation risk was manipulated by broadcasting playbacks of either predator- or non-threatening- calls and sounds, every few minutes, throughout the 130 day breeding season, over a 16 ha area. A given female was exposed to either predator playbacks throughout the season (predator playback females) or non-threatening playbacks. We assessed the effects on nest site selection, skittishness (flight initiation distance, FID), the duration of incubation bouts, and the number of feeding bouts per hour during brood-rearing. We could be certain any resulting effects on demography were attributable to predation risk alone because we successfully eliminated direct predation by protecting every nest using electric fencing and seine netting.
Results/Conclusions
Predator playback females located their nests in denser, thornier vegetation, and nesting in denser, thornier vegetation affected incubation behaviour, and was associated with poorer nestling condition and a greater proportion of nestlings expiring prior to fledging. Predator playback females were more skittish (> FID), and greater skittishness was associated with lower hatching success, a lower feeding rate during brood-rearing, poorer nestling condition, and a greater proportion of nestlings expiring prior to fledging. Predator playback females spent shorter times on and longer times off the nest during incubation, which a previous study showed is associated with poorer hatching success. Finally, predator playback parents made fewer feeding visits per hour during brood-rearing, which was associated with poorer nestling condition and a greater proportion of nestlings expiring prior to fledging. The cumulative cost of these anti-predator behaviours was dramatic as predator playback females fledged almost 40 % fewer offspring over the course of the breeding season than those exposed to non-threatening playbacks (3.8 ± 0.4 vs. 6.0 ± 0.4, mean ± SE, P < 0.001). We conclude that our experimental results provide clear evidence that the costs of anti-predator behavioural responses may significantly affect the demography of wildlife populations.