LAUSR

Insect epidemics such as the mountain pine beetle (MPB) outbreak have a major impact on forest dynamics. In Cypress Hills, Canada, the Forest Service Branch of the Saskatchewan Ministry of Environment aims to control as many new infested trees as possible by conducting ground-based surveys around trees infested in previous years. Given the risk posed by MPB, there is a need to evaluate how well such a control strategy performs. Therefore, the goal of this study is to assess the current detection strategy compared with competing strategies (random search and search based on model predictions via machine learning), while taking management costs into account. Our model predictions via machine learning used a generalized boosted classification tree to predict locations of new infestations from ecological and environmental variables. We then ran virtual experiments to determine control efficiency under the three detection strategies. The classification tree predicts new infested locations with great accuracy (AUC = 0.93). Using model predictions for survey locations gives the highest control efficiency for larger survey areas. Overall, the current detection strategy performs well but control could be more efficient and cost-effective by increasing the survey area, as well as adding locations given by model predictions.