Implications of swede midge biology in the development of population management alternatives

The swede midge, Contarinia nasturtii, is a significant pest of cruciferous crops in Canada and the United States. Though some control has been achieved with chemical pesticides, there is a need to evaluate alternative tactics for organic producers. The purpose of this study is to examine how C. nasturtii life history traits impact the potential success of alternative control tactics. The effect of temperature on three entomopathogenic nematode species, and one entomopathogenic fungus were investigated. Successful reproduction and infectivity were observed in all nematode species and a nematode-fungus combination generally caused highest mortality to all C. nasturtii life stages. The fungus suppressed emergence from soil in 2012 and all nematode treatments suppressed emergence in 2013 in broccoli fields. Biopesticides were evaluated for larval mortality and oviposition deterrence in the greenhouse, and damage symptoms in the field. In greenhouse trials, treatments resulted in significant reductions, with a greater proportion of reductions attributable to oviposition deterrence than larvicidal activity. Field trials showed the potential for biopesticides to reduce feeding damage; however, efficacy was dependent on C. nasturtii abundance. Insect exclusion fencing and planting dates were evaluated as tactics to reduce C. nasturtii damage through spatial and temporal avoidance. Fencing increased the number of marketable plants when plots were separated from fields infested with C. nasturtii by ~800 m, but no effect was observed with lower spatial separation. A survey of flight height produced observations of adults flying 240 cm above the soil, which is above the height of the fencing. Reduced damage was observed in late broccoli plantings and early cabbage plantings, while traditional planting dates suffered total yield losses. The population model MidgEmerge was evaluated for prediction accuracy of spring emergence, and sensitivity to temperature, rainfall, and model parameters. Predictions were early during hot, dry years, and late during cool, wet years. Temperature affected model output, but rainfall did not. Manipulating thresholds for ‘diapause larva development temperature’, and ‘minimum temperature for pupal growth’ resulted in changes to predicted emergence. Prediction by MidgEmerge potentially overstates the effect of temperature, and understates the effect of rainfall. Recommendations to improve model accuracy are provided.