This three-year project was completed in mid-2022.
When a new insect herbivore species arrives in NZ it is uncertain whether it will be able to establish in natural ecosystems and feed on native plants. This project investigated whether non-native species share traits which might enable DoC and MPI risk analysts to predict whether or not the species might pose a risk in natural ecosystems.
A database of a wide range of taxonomic, biological, ecological and distribution data populated with attributes of weevils, and later, aphids was developed. Analysis showed that non-native species that feed on NZ native plants more often have other taxonomically related species present in NZ; occur in Australia; have natural distributions with climatic similarities to NZ and tend towards polyphagy.
A Random Forest model using attributes for 65 non-native weevil species correctly predicted 46 of 65 observations (71%) in separating non-native weevils that feed on NZ native plants from those that do not. Analysis of the aphid data for 134 non-native species established in NZ showed that the subfamily Aphidinae are particularly polyphagous and more likely to feed on NZ plants; are particularly likely to occur in “pristine” NZ habitats and are larger than those recorded on NZ plants in less pristine habitats. We concluded that there are attributes of non-native herbivores in these taxa that could help predict whether native plants could be at risk from new incursions.
Implementation: The completed framework will be used by DOC and MPI Science & Risk – Plants & Pathways, as part of their risk assessment processes, to identify biosecurity threats to New Zealand’s natural systems.
Zotero can be found under ‘Outputs’ on the B3 homepage.