Dispersal and competitive release affect the management of native and invasive tephritid fruit flies in large and smallholder farms in Ethiopia

African horticulture is seriously affected by fruit flies, both native and invasive. Novel sustainable control methods need testing against the backdrop of smallholder-dominated farming of Africa. We evaluated the potential of male-specific attractants (parapheromones) laced with insecticide to suppress the alien invasive Bactrocera dorsalis and native Ceratitis capitata. In large-scale guava, methyl-eugenol (ME)-bait stations combined with toxic protein baits suppressed B. dorsalis within 8 months but resulted in a resurgence of the displaced Ceratitis capitata. In smallholder farms, intervention using SPLAT-ME laced with spinosad was surprisingly unsuccessful. Subsequent mark-release-recapture experiments showed high dispersal rates of flies, covering many times a typical farm size, leading to a continuous influx of flies from surrounding areas. Several other factors important for intervention were evaluated. SPLAT-MAT-ME dollops remained attractive for over two weeks, although gradually becoming less attractive than fresh baits. Further, competitive displacement was observed: C. capitata selectively emerged from fruits in which B. dorsalis infestation was low. Finally, we evaluated whether ME could be combined with C. capitata male attractants [trimedlure (TML) and terpinyl acetate (TA)] without affecting attraction. Combining male lures did not affect catches directly, although at very high populations of B. dorsalis attracted to ME interfered with C. capitata trap entry. Although ME-based methods can effectively suppress B. dorsalis, they were not effective at single smallholder scale due to the high dispersive propensity of tephritids. Further, competitive release implies the need for a combination of lures and methods. These observations are important for developing control schemes tailored for African smallholder settings.