Combining surveillance data from, and comparing the relative utility of wildlife sentinels for detecting bovine tuberculosis in New Zealand

Brushtail possums (Trichosurus vulpecula) are the main wildlife host of bovine tuberculosis in New Zealand. The disease cycle can be broken by intensive large-scale reduction of possum populations to low levels. That strategy has been successfully applied over much of New Zealand, to the point that a key question now is whether the disease has indeed been locally eliminated, so that the expensive disease control programme can be stopped. However, it is usually too expensive to survey the few possums left. We have therefore developed a new approach to characterising the Tb detection capabilities of any wild animal species that can host the disease. The approach assigns each individual animal and spatially-explicit “detection kernel” that combines data on average range utilisation with probabilities of becoming infected when Tb is present in possums at the lowest sustainable level. We describe the principles involved, and compare the detection capabilities and cost-effectiveness across the species that are commonly being used for this purpose – wild pigs (Sus scrofa), ferrets (Mustela furo), and deer (Cervus spp.).