# Eradication of Yellow Crazy Ants, Anoplolepis gracileps Smith, from Lismore and Statistical Proof of Freedom Using Scenario Tree Analysis

## TL;DR

This paper describes the eradication of yellow crazy ants in Lismore, Australia, and uses statistical analysis to determine when eradication efforts can safely stop.

## Contribution

The paper introduces scenario tree analysis to provide statistical proof of eradication success for invasive species.

## Key findings

- Scenario tree analysis predicted a 70.4% probability of freedom for one detected nest.
- The probability of freedom increased to 98% for five nests using the same analysis.
- The method helped inform when to terminate eradication and surveillance activities.

## Abstract

Exotic insects frequently invade countries, and many countries attempt to eradicate these exotic incursions. Regulatory authorities continue to monitor infested areas to understand whether the eradication methods were working and, finally, whether the eradication was successful. A common problem is how long should eradication and monitoring continue after the last invader is detected, because these are costly activities. We use the example of yellow crazy ant detection and eradication. Here, we used the statistical analysis of probabilities using scenario tree analysis to predict the likelihood of eradication based on different assumptions. This methodology has been used in both plant and animal kingdoms and will assist decision makers regarding when to cease activities.

Yellow crazy ants (YCAs) are an invasive ant with a pantropical distribution, largely due to the international movements of ships and produce. This invasive ant has the capacity to impact a broad range of environmental, domestic and agricultural situations and has the ability to develop into supercolonies and dominate landscapes if uncontrolled. YCAs have been detected in several locations in Australia. During 2018 in New South Wales, YCAs were detected in two locations in the Lismore region. Several awareness techniques were used to gain community support and engagement in the response program. The eradication program relied on the insecticide fipronil (several formulations), and the program subsequently used surveillance data to demonstrate that eradication had been achieved. We used the scenario tree analysis with stochastic models to estimate the likelihood of eradication. We combined the results of the passive and active surveillance systems to predict a 70.4% (62.7–80.7) probability of freedom of detecting one nest, 84.4% (73.9–94.4) probability of freedom for two nests and 98% (93.1–99.9) probability of freedom for five nests. The results from the scenario tree analysis were used to inform program managers regarding the termination of the eradication and surveillance activities.