An index for selective toxicity of miticides to phytophagous mites and their predators based on orchard trials.

Data from nine trials conducted from 1990 to 1998 in apple orchards in Nova Scotia and Quebec, Canada, were used to estimate the predator-prey selectivity of miticides and their potential compatibility with biological control of mites. The European red mite Panonychus ulmi (Koch) was the dominant and more harmful phytophagous species, followed by the apple rust mite, Aculus schlechtendali (Nalepa). Two predacious mites, the phytoseiid, Typhlodromus pyri Sheuten, and the stigmaeid, Zetzellia mali (Ewing), were often found in the orchards. We used one minus the ratio of mite-days in treated plots to those in the control plots as an index of population suppression and toxicity of the miticides. Miticides were then categorized into classes similar to those employed by the International Organization for Biological Control to rate pesticide toxicity to natural enemies of insect and mite pests. Selectivity of miticides was mostly based on toxicity to P ulmi, the major pest, versus toxicity to T pyri, the major predator, with some consideration of the two lesser species, A schlechtendali and Z mali. In most cases, our findings were in accord with other studies. Abamectin and clofentezine had favourable selectivity (more toxic to the two phytophagous mites than to T pyri). The higher recommended rate of pyridaben (450 g ha(-1)) and two rates of spirodiclofen (180 and 240 g ha(-1)) were neutral (equally toxic to pests and predators). The lower rate of pyridaben (216 g ha(-1)), dicofol, formetanate hydrochloride and propargite were unfavourably selective (more toxic to T pyri). A higher than recommended rate of pyridaben (2160 g ha(-1)) applied before bloom was disruptive--P ulmi-days after treatment were actually greater than with the untreated control. P ulmi resistance to dicofol and propargite were probable complicating factors in some of the orchard trials.