Laboratory tests to estimate the non-target impacts of four Aphidius spp. parasitoids in the field

Abstract

A ‘risk score’ approach to help determine the potential impact of biological control agents on non-target species, previously proposed for weed biocontrol agents, was examined in an ex post study for insect biocontrol agents. ‘Risk scores’ (% parasitism for non-target aphid species/% parasitism for target aphid species) were calculated for 19 aphid (New Zealand adventive and indigenous) species and 4 introduced (New Zealand) Aphidius biocontrol species, based on previously published choice and no-choice host range laboratory testing. Field hosts for each Aphidius parasitoid and aphid host combination were also examined through an extensive literature review. Strong correlations between ‘risk scores’ for both choice and no-choice laboratory assays and field host binary values (host: yes or no) were obtained for each Aphidius/aphid dataset, but there was no threshold score that indicated using the approach to predict parasitoid field host ranges could result in type II errors. In contrast, a second analysis that categorized aphid hosts as ‘major field hosts’, if literature records indicated high (≥15%) levels of parasitism in the field, showed a clear threshold risk score predicting whether an aphid was a major field host in no-choice tests. We conclude that this approach could potentially be used to predict the likelihood of the impact of Aphidius biocontrol agents on non-target aphid species. We discuss the pros and cons of this approach to help scientists and regulators in the assessment for new biocontrol agent introductions, including the need for this ‘risk score’ approach to be tested on a greater range of insects.