Search

A species in the Bactrocera dorsalis (Hendel) complex was detected in Kenya during 2003 and classified as Bactrocera invadens Drew, Tsuruta & White. Having spread rapidly throughout Africa, it threatens agriculture due to crop damage and loss of market access. In a recent revision of the B. dorsalis complex, B. invadens was incorporated into the species B. dorsalis. The potential distribution of B. dorsalis has been previously modelled. However, previous models were based on presence data and did not incorporate information on the seasonal phenology of B. dorsalis, nor on the possible influence that irrigation may have on its distribution. Methyl eugenol-baited traps were used to collect B. dorsalis in Africa. Seasonal phenology data, measured as fly abundance throughout the year, was related to each location's climate to infer climatic growth response parameters. These functions were used along with African distribution records and development studies to fit the niche model for B. dorsalis, using independent global distribution records outside Africa for model validation. Areas at greatest risk of invasion by B. dorsalis are South and Central America, Mexico, southernmost USA, parts of the Mediterranean coast, parts of Southern and Eastern Australia and New Zealand's North Island. Under irrigation, most of Africa and Australia appear climatically suitable. [ABSTRACT FROM PUBLISHER]

Two correlative approaches to the challenge of ecological niche modeling (genetic algorithm, maximum entropy) were used to estimate the potential global distribution of the invasive fruit fly, Bactrocera invadens, based on associations between known occurrence records and a set of environmental predictor variables. The two models yielded similar estimates, largely corresponding to Equatorial climate classes with high levels of precipitation. The maximum entropy approach was somewhat more conservative in its evaluation of suitability, depending on thresholds for presence/absence that are selected, largely excluding areas with distinct dry seasons; the genetic algorithm models, in contrast, indicate that climate class as partly suitable. Predictive tests based on independent distributional data indicate that model predictions are quite robust. Field observations in Benin and Tanzania confirm relationships between seasonal occurrences of this species and humidity and temperature. [ABSTRACT FROM AUTHOR]