PS 9-103 - Influence of vegetation density on the transmission and infection severity of coffee leaf rust in shaded coffee agriculture

Monday, August 7, 2017
Exhibit Hall, Oregon Convention Center
Minh Chau N. Ho, Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, MI and John H. Vandermeer, Department of Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, MI
Background/Question/Methods

Coffee leaf rust is a common agricultural disease in coffee plantations and is caused by the fungal pathogen Hemileia vastatrix. In 2013, epidemics throughout Latin America caused devastating economic loss and re-invigorated research into new disease management. On some coffee plantations, trees are used for ecosystem services such as nitrogen fixation, fruits, and to support native biodiversity. Because H. vastatrixuses wind for transmission and humidity for germination, we propose that coffee leaf rust disease dynamics can be influenced by tree stands. Specifically, we hypothesize that tree stands will disrupt wind transmission of fungal spores, but that increased canopy cover will reduce evaporation, increase local humidity, and increase rust germination. In 2016, we explored the effect of trees on local disease dynamics on a highland coffee plantation in Chiapas, Mexico. At 128 sites, we measured evaporation rates, tree density, coffee plant density, and canopy cover, as well as the number of infected leaves and the total number of leaves on five coffee plants. We predicted that increased tree density will reduce the probability of coffee plants contracting infection, but that once infected, coffee plants under higher canopy cover will experience less evaporation and suffer higher infection intensity.

Results/Conclusions

We modeled the influence of tree density, canopy cover, evaporation, and coffee plant density on the number of infected leaves per plant using a general linear mixed model for a negative binomial distribution with zero-inflation. Coffee plants were less likely to become infected at higher tree and coffee plant densities. However, increasing canopy cover increased infection incidence. The number of infected leaves was influenced only by higher coffee densities, and evaporation had no influence on infection. This suggests that structural variables – such as trees and coffee plants themselves – block wind-dispersal of rust spores and reduce the probability of plants contracting the coffee leaf rust. However, once infected, the intensity of the infection is controlled by higher coffee plant densities in the vicinity, not humidity. We suspect that areas with higher coffee densities are more likely to contain different varieties, some of which are more resistant than others, so that disease severity is managed through the dilution effect. These results suggest that tree stands can provide additional ecosystem services in controlling the coffee leaf rust disease, though they will need to be used in conjunction with other disease management strategies.