Site_Detect: a design tool for block-based pest surveillance

Growers maintain a range of practices to manage pests on their properties, but even with good management in place, pests can enter farms from neighbouring areas – particularly if there are abandoned farms, wild fruit trees and urban areas nearby.

Insect trapping is a common practice on horticulture farms to monitor the presence of insect pests. Surveillance traps are particularly important for monitoring pests of ‘quarantine concern’ for target markets.

Monitoring at the farm (or farm block) scale can be used in addition to landscape scale surveillance, or where there is no area-wide monitoring program in place. The advantage of farm-level monitoring is that this is the scale where business decisions are made, and where growers have greatest control of management responses.

The challenge for growers is how to place traps around their property to give the best chance of detecting pests already present on site as well as pests that may be entering from outside.

The Site_Detect model shows how surveillance design (trap density and arrangement), pest dispersal ability and lure attractiveness affects the probability of detecting pests at the site scale.

Site_Detect is now available as a shiny app, allowing biosecurity stakeholders to test multiple scenarios for trapping designs.

What you can do with the Site_Detect tool

Users can select various trapping arrangements, lure attractiveness, and different characteristics of a pest outbreak. It shows the probability of detection for pests entering the site from outside versus those already present inside the site, and the effects of trap location. The app can help industry and biosecurity regulators find a trapping design that achieves a balance between being practical to implement and sufficiently sensitive to detect pests.

What we learned from the model

The model has been peer reviewed and published in the Journal of Economic Entomology here:

Trap arrangement

The model simulates pest detection with multiple scenarios looking at the size and shape of the site, traps (arrangement, number and attractiveness) and pests (origin point, number and spread).

We compared the ability of three trap arrangements (regular, random, and perimeter) to detect pests. We found that a regular arrangement was best overall for detecting pests already in the site as it gave superior site coverage. Perimeter traps were best for detecting pests entering the site. Interestingly, randomly placed traps performed relatively well compared to regularly placed traps, which suggests that the exact placement of traps isn’t so important.

Lure attractiveness and trapability

Lure attractiveness and trap density turned out to be the most influential factors affecting pest detection probability. High lure attractiveness combined with high trap density increases the probability of immediately detecting a pest, even if traps are not optimally placed.