RESUMO
The tea green leaf hopper, Empoasca onukii Matsuda, is a severe pest of tea plants. Volatile emissions from tea shoots infested by the tea green leafhopper may directly repel insect feeding or attract natural enemies. Many studies have been conducted on various aspects of the tritrophic relationship involving tea plants, tea green leafhoppers and natural enemies. However, mathematic models which could explain the dynamic mechanisms of this tritrophic interaction are still lacking. In the current work, we constructed a realistic and stochastic model with temperature-dependent features to characterize the tritrophic interactions in the tea agroecosystem. Model outputs showed that two leafhopper outbreaks occur in a year, with their features being consistent with field observations. Simulations showed that daily average effective accumulated temperature (EAT) might be an important metric for outbreak prediction. We also showed that application of slow-releasing semiochemicals, as either repellents or attractants, may be highly efficacious for pest biocontrol and can significantly increase tea yields. Furthermore, the start date of applying semiochemicals can be optimized to effectively increase tea yields. The current model qualitatively characterizes key features of the tritrophic interactions and provides critical insight into pest control in tea ecosystems.