Beyond Borders: Managing Aquatic Invasive Species With Community Detection Algorithms
Aquatic invasive species (AIS) pose a dire threat to Minnesota’s freshwater ecosystems. Due to the difficulties encountered when attempting to manage AIS through selective pesticide or physical removal, invasion prevention is often recommended in favor of remediation. The movement of recreational watercraft throughout Minnesota can inform the dispersal of prevention efforts. AIS Explorer is an online decision-support tool that uses a network model to describe the watercraft movements and in turn assess the risk of AIS invasion across the state’s waterways. Our work sought to utilize community-detection algorithms to identify subsections of these watercraft movement networks with elevated invasion risk and more precisely describe potential patterns of invasive species spread throughout the state. In this study, we examined the lakes of Crow Wing county using these community-detection algorithms. We identified a subset of lakes from each of the detected community clusters and used a particle swarm optimization algorithm to rank the candidate lakes based on their position within the original boater movement network. We used these scenarios to inform interventions and simulated these interventions using AIS Explorer’s predictive spread model. Intervention efficacy was evaluated based on the number of novel AIS infestations occurring after intervention in each scenario. Results from these simulations could be used in conjunction with the knowledge of local natural resource managers to tailor interventions to the local ecosystem.