Reducing the primary exposure risk of Henderson crakes (Zapornia atra) during aerial broadcast eradication by selecting appropriate bait colour

Abstract Context. Operations to eradicate non-native invasive predators from islands frequently put native species at risk of consuming harmful substances, such as poison bait. The incorporation of certain colours in poison-bait pellets may reduce the risk of bait consumption and, therefore, non-target mortality. Previous work indicated that birds generally avoid blue or green colours; however, there is substantial inter-specific variation in this preference, and more experimental work on species of conservation concern is needed. Aims. We tested whether a globally threatened island endemic, the Henderson crake (Zapornia atra), which suffered substantial mortality during a rat-eradication attempt on Henderson Island in 2011, would consume fewer blue than green pellets, which were used during the previous eradication attempt. Methods. We held 22 Henderson crakes in captivity and provided them with either blue or green non-toxic pellets for 5 days in June and July 2015. We measured consumption and used linear mixed models to evaluate whether bait colour influenced consumption. Key results. Henderson crakes did not consume any dry pellets, and all trials were conducted with wet bait pellets. We found slightly lower consumption of blue pellets than green pellets, and substantial variation among individuals. Females (n = 17) consumed 24% less blue than green bait, whereas males (n = 5) consumed 77% less blue than green bait. Conclusion. Henderson crakes are unlikely to consume dry pellets, and will likely consume fewer blue than green bait pellets. Implications. We recommend that any future rat eradication on Henderson Island considers using blue rather than green baits and targets dry weather to reduce the risk of Henderson crakes consuming toxic rodenticide bait pellets.