The indirect effects of a keystone ant-hemipteran mutualism on coffee berry load

Background/Question/Methods

Ant-Hemipteran interactions are common and can benefit the host plant if the negative effect of the Hemipteran is offset by the deterrence of more voracious herbivores by ants. Ant-Hemipteran interactions are important in agricultural systems as the outcome of the interaction affects crop yield. The green coffee scale (Coccus viridis) is a damaging herbivore of coffee and citrus. In Mexican coffee farms, its ant mutualist, Azteca instabilis, allows C. viridis populations to become dense, and subsequently infected by Lecanicillium lecanii, a fungus that hyperparasitizes the coffee rust fungus (Hemileia vastatrix). The goal of this study was to determine if this ant-Hemipteran mutualism has an overall benefits coffee berry load because it indirectly lowers the incidence of a plant pathogen. Surveys were conducted on an organic coffee agroecosystem in Chiapas, Mexico. The number of healthy/infected scale insects, coffee rust lesions, and ant activity were documented monthly during 2009. In 2010, the height, amount of mature coffee berries, number of reproductive branches, variety and shade cover were recorded for each coffee stem. A zero-inflated negative binomial (ZINB) Bayesian hierarchical model was used to determine the effects of the covariates on coffee berry load. 

Results/Conclusions

We could evaluate the effect of H. vastatrix on berry load because it does not follow the plant vigor hypothesis. The same is true for dense C. viridis populations: their distributions are determined by ants, which restrict foraging to coffee bushes within ~3 meters of their arboreal nest, and shade had no effect on yield. Berry load in 2010 was decreased by late season H. vastatrix in 2009. Greater C. viridis densities in 2009 tended to increase berry load, but not significantly. Previous studies have found that C. viridis infected by L. lecanii will reduce the prevalence of H. vastatrix the next year, so this weak effect of C. viridis on production may be due to lower H. vastatrix. Further, C. viridis infected by L. lecanii in 2009 were negatively correlated with the prevalence of H. vastatrix in 2010, so we could predict that yield in 2011 will be higher in these areas. While the observed Azteca - C. viridis complex does not have significant negative effects on coffee berry load, C. viridis is well-documented in the literature to decrease yield at extreme densities; it could be that the observed C. viridis densities are not yet harmful for the plant.