Predicting Within- and Between-Year Variation in Activity of the Invasive Spotted Wing Drosophila (Diptera: Drosophilidae) in a Temperate Region

Abstract

Invasive insect pests can be challenging to manage because their recent arrival provides limited information on which to build predictive population models. The magnitude and timing of activity by the invasive vinegar fly, Drosophila suzukii, in crop fields has been unpredictable due to its recent arrival in many new regions of the world and changes in methods for its detection. Using 7 yr of consistent trapping of adults at four blueberry farms in Michigan, United States, we modeled the temporal and environmental factors influencing D. suzukii activity. We found that this pest established high levels within 2 yr of being detected, with peak fly activity continuing to increase. Fly activity timing and abundance were predicted by the annual number of days below 0°C, the number of winter and spring days above 10°C, and by the fly activity in the preceding year, providing support for overwintering in our region. We monitored larval infestation for 4 yr at these same sites and found a moderate positive correlation between larvae in fruit and adults in traps. Finally, we developed a generalized additive model to predict D. suzukii fly capture throughout the season based on relevant environmental factors and examined the relative timing and magnitude of activity under varying winter and spring temperature conditions. Our results suggest that D. suzukii activity is predictable and that environmental conditions can be used in temperate regions to provide regional risk warnings as a component of strategies to manage this invasive insect pest.

Keywords: GAM; establishment; generalized additive models; invasive species; predictive modeling.