Establishment of viable populations of biological control agents is critical in weed biological control programs. Existing native predators pose a substantial, often overlooked risk to agent establishment and success. Native predators have the potential to exert negative direct effects on biological control agent populations with indirect positive effects on the plant populations that the agents were introduced to control (i.e., a species cascade). Quantification of these effects, in addition to understanding predator distribution and abundance, is essential to design of effective biological control releases that minimize the potential risk of failure due to predation.

Results/Conclusions

We identified predation of Tyria jacobaea (Cinnabar moth) larvae by native Camponotus sp. (carpenter ants) on the invasive weed, Senecio jacobaea (tansy ragwort). Camponotus sp. colony density was significantly higher in drier areas with a more recent history of disturbance (<12 years) by wildfire and logging than in moist areas with a less recent history of disturbance by wildfire and logging (>12 years). Initial observations revealed fewer remaining T. jacobaea larvae in plots where Camponotus sp. were present. Further experimentation revealed that more T. jacobaea larvae were taken from plants accessible to ants than those stems where ants were excluded. In addition, more larvae were taken in the dry burned and logged areas overall, than in the moist, less recently disturbed areas. In a final experiment in the dry, disturbed area the relative percent of buds or flowers consumed by T. jacobaea larvae was significantly higher on plants where larvae were protected from Camponotus sp. compared to those that were accessible to Camponotus sp. showing strong evidence of a species cascade in this system. The results of this study emphasize the importance of not only identifying potential predators to biological control agents and quantifying their effects at the level of both the agent and the invasive plant, but also the environmental context in which they may pose the greatest risk to the establishment and subsequent efficacy of a particular biological control agent.