Fruit flies are a serious pest of horticultural industries throughout the world. The sterile insect technique is used in many countries to suppress or eradicate wild populations. The technique would be more effective if there was a better understanding of the interaction between wild and sterile fruit flies, and their interaction with their environment. We reviewed historic data from the 1998/99 sterile release of Queensland fruit fly (Qfly) at Wagga Wagga, New South Wales which has a resident population of wild Qfly. Sterile flies were released at one point and the wild and sterile numbers were analysed in relation to each other, and to climate and landscape variables. Fifty seven fruit fly traps on a 400 m array within 7 km were used to collect the data. Regression analysis, accounting for 20.1% of the variance, indicated that log number wild Qfly was best predicted by the model with the factors of altitude, growth index, day length, cold stress, rainfall, minimum temperature and dry stress. Log number wild Qfly increased as day length and minimum temperature indices increased but was negatively related to other variables. Regression analysis, accounting for 27.2% of the variance, indicated that log sterile Qfly number was best predicted by a model with parameters including distance from release point, moisture index and maximum temperature. All parameters were inversely related to log number sterile Qfly. Fletcher's inverse distance square rule was tested. The best model linked log number sterile Qfly (including zero detections) with log distance and accounted for 22.5% of the variance. Of the 19 758 sterile Qfly trapped, 50%, 95%, 98% and 100% were trapped within 556 m, 1 355 m, 1 574 m and 5 553 m respectively of the release site. The mean dispersal distance of sterile flies was 830 m and this was used to estimate quarantine distance for wild flies.
|Number of pages||6|
|Journal||Plant Protection Quarterly|
|Publication status||Published - 2013|
- flight distance
- Beta distribution
- predictive models