The known unknowns in international border interceptions of non-native insects

Abstract Propagule pressure is one of the key drivers of establishment success of non-native species, including insects. However, border interception data, which have the potential to act as a proxy for true arrival rates (i.e., propagule pressure) of insects into a country, are seldom used to predict invasions. This can be due to the limited amount of interception data in some countries, difficulties accessing such data, and when these data are available, difficulties in addressing biases caused by variation in interception probability for different taxa due to policy changes, and operational influences of import inspections. The type of interception data required to reliably estimate arrival rates is rarely available. To improve the use of interception data as a proxy for propagule pressure, we investigated the fraction of established species which had interceptions and vice versa by taxonomic group and by biological characteristics (development type and feeding group), using several national datasets from five continents. We identified higher fractions of established species that were intercepted for plant feeding insect groups compared to fractions of non-plant feeding groups, even in countries with more general import inspection strategies. This is likely to reflect greater search effort for and recording of plant-feeding insect groups during inspections. To address this variation in interception probability and improve establishment predictions based on interception frequency, we developed a methodology to account for such taxonomic variation. We apply this to three hazard lists of insects, containing potential pests for an industry or potential pests for a country, to assess its effectiveness.