Incipient founder populations of Mediterranean and Queensland fruit flies in Australia: the relation of trap catch to infestation radius and models for quarantine radius

Abstract

We examined data from 75 infestations of the Mediterranean fruit fly (Medfly) and 286 of the Queensland fruit fly (Qfly) that have occurred in quarantined and normally fly-free zones in Australia from 1974 to 2000. The radius of occurrence of both adult male flies and infested fruit was almost always less than 1 km. The rare cases where there was an isolated occurrence beyond 1 km of an epicentre were most likely due to (and can be treated as) separate introductions. Our analysis shows that effective quarantine radii for suspension of fly-free status should be related to the number of flies trapped around the epicentre and the density of the trap array (if the appropriate code of practice is applied). Most detections of fruit flies involve the trapping of very few flies and 18% of Medfly infestations and 71% of Qfly infestations that are detected are not classified as outbreaks and are left to die out without any treatment. For each species, we have used 3 alternative methods to calculate confidence limits for infestation radii. The upper limits could also serve as quarantine radii. These limits increase with the rate of trapping of male flies and have a theoretical probability of 3/100 000 (i.e. probit 9) of being exceeded. The quarantine radii for most declared outbreaks, when calculated with any of our methods, would be small because the number of flies detected is usually only just above the threshold for such a declaration. If our methods were used for beneficial species or for re-introductions of endangered species, the lower confidence limits could be used to calculate the size of inoculum required for a high probability of initial establishment.