Potential for autonomous control of the coffee rust disease from the join action of two control agents

Background/Question/Methods

Recent theoretical work by Ong and Vandermeer (2015), has demonstrated the importance of considering the complexities of natural ecosystems in autonomous pest control. They were able to demonstrate increased effectiveness of incorporating both a pathogen and predator as simultaneous control agents for a pest, in contrast to the one-dimensional approach of only pathogen or predator alone. The coffee rust, caused by the fungi Hemileia vastatrix, serves as a convenient model system to test this theoretical framework. The coffee rust has a mycoparasitic fungal pathogen, Lecanicillium lecanii, and a less noted small dipteran larva predator, Mycodiplosis sp., which is known to be a voracious predator on rust spores. Neither L. lecanii nor Mycodiposis alone have been successful biological control agents according to most reports, and their potential, as an integral part of an autonomous system of control has not been investigated. To investigate the effectiveness of L. lecanii and Mycodiplosis sp. in controlling the coffee rust, comparative surveys were conducted in December 2014 on a coffee farm in the Soconusco region of Chiapas, Mexico, where the coffee rust regularly reaches high densities, and on a coffee farm near Orocovis in Puerto Rico where coffee rust is only found at low densities. 

Results/Conclusions

We found a statistically significant correlation (GLM, p<0.05) between per-plant rust infection and larval density, with only highly infected areas containing significant numbers of the dipteran larvae and suggesting some density dependence effect. In Puerto Rico we observed 88% of coffee leaves were infected with L. lecanii and 68% of coffee leaves had Mycodiplosis sp. larvae with an average number of 2.8 Mycodiplosis sp. larvae per leaf. In Mexico we observed 55% of the coffee leaves were infected with L. lecanii and 32% of the coffee leaves had Mycodiplosis sp. larvae with an average number of 0.9 Mycodiplosis sp. larvae per leaf. The pattern of higher incidence of both predator (Mycodiplosis sp. larvae) and pathogen (L. lecanii) in Puerto Rico, where the coffee rust does not reach high densities, supports the hypothesis that these two separately unstable control agents, when combined, may be providing stable (or persistent) autonomous pest control in Puerto Rico and not Mexico.