Abstract
Rodent predators are implicated in declines of seabird populations, and removing introduced rats, often, but not always, results in the expected conservation gains. Here we investigated the relationship between small mammal (Norway rat, wood mouse and pygmy shrew) abundance and Manx shearwater breeding success on the island of Rum, Scotland, and tested whether localised rodenticide treatments (to control introduced Norway rats) increased Manx shearwater breeding success. We found that Manx shearwater breeding success was negatively correlated with late summer indices of abundance for rats and mice, but not shrews. On its own, rat activity was a poor predictor of Manx shearwater breeding success. Rat activity increased during the shearwater breeding season in untreated areas but was supressed in areas treated with rodenticides. Levels of mouse (and shrew) activity increased in areas treated with rodenticides (likely in response to lower levels of rat activity) and Manx shearwater breeding success was unchanged in treated areas (p < 0.1). The results suggest that, unexpectedly, negative effects from wood mice can substitute those of Norway rats and that both species contributed to negative impacts on Manx shearwaters. Impacts were intermittent however, and further research is needed to characterise rodent population trends and assess the long-term risks to this seabird colony. The results have implications for conservation practitioners planning rat control programmes on islands where multiple rodent species are present.