Estimating non-target effects to better assess the risk of biological weed control

Non-target effects are a major risk in implementing any weed biological control project which are currently minimized through extensive pre-release host specificity testing. However, after a biological control agent has been approved and released, non-target effects of biological control are seldom adequately assessed at the population level. This research addresses this gap in knowledge by presenting a framework to estimate non-target effects of introduced biological control agent the cinnabar moth, Tyria jacobaeae L. on a non-target host arrowleaf groundsel, Senecio triangularis Hook. 

Ecological risk of this non-target interaction is being assessed through a standard risk = hazard * exposure model across spatial and temporal scales. Hazard is being determined through an experiment manipulating both the timing and severity of herbivory by cinnabar moth larvae on S. triangularis. Hazard will then be quantified as the sensitivity of S. triangularis growth, reproduction, and survival to the range of herbivory treatments experienced. The sensitivity to herbivory experienced in the manipulative experiment is linked to exposure through observations of insect-plant interactions collected at 24 S. triangularis sites across Western Oregon between years 2003 and 2016. Together, the experiment and regional exposure survey document and assess risk of population level non-target effects across changing Pacific Northwest environments. Ultimately this research will provide further evidence about the risks and benefits of biological control to better inform decisions by the public, land managers, and policy makers.