Ex ante analyses can predict natural enemy efficacy in biological control

Abstract

The success rate of biological control programs against invasive species is low because the efficacy of released natural enemies is unknown. Deconstruction of the successful biological control of the highly destructive cassava mealybug (CM) and cassava green mite (CGM) in Africa is used to show how ex ante pre-release analyses of natural enemies can increase the success rate. A meta-population tri-trophic physiologically based demographic model (PBDM) of the cassava system accurately predicted the weather-driven daily dynamics of cassava, CM, CGM, and their introduced natural enemies across the ecological zones of Africa. Marginal analyses of the simulation data enabled parsing the differing contributions of two exotic parasitoids and a fungal pathogen in the control of CM, and the contributions of two mite predators, local pathogens and rainfall mortality in the control of CGM. The contribution of each natural enemy to the [~]95% recovery of cassava yields is estimated. The results show ex ante analyses of prospective natural enemies using well parameterized PBDMs can predict natural enemy prerelease efficacy across vast geographic areas and increase the success rate of biological control programs. Ex ante analyses can also explain whether biological control is a viable option (e.g., Cure et al. 2020).