Fungal Biocontrol in Coffee: A Case Study in Agroecology.

Abstract

In light of mounting concerns over the profound environmental, health, and social costs of conventional agriculture, there is a compelling need to develop alternatives for growing food and fiber. Development of a truly sustainable agriculture will require a forward-looking agenda that embraces the tools of modern science. However, what is needed is not the reductionist science of the Green Revolution, but rather a science that recognizes and embraces the inherent complexity of natural systems, i.e., an agricultural science that benefits from and contributes to the science of ecology. This dissertation details a case study in which this paradigm, agroecology, was applied to the study of a fungal biocontrol agent in a coffee agroecosystem. The fungus Lecanicillium lecanii attacks two potential coffee pests: the green coffee scale (Coccus viridis) and coffee rust (Hemileia vastatrix). Despite the potential importance of this biological control agent, there is still much that is unknown about its basic ecology. The first half of this dissertation represents a step toward addressing this deficiency. Through a combination of field and laboratory studies, it was shown that L. lecanii can persist in the soil; that it can be translocated from the soil via rain splash; and that the ant Azteca instabilis can subsequently spread the fungus between individuals of C. viridis. The potential for L. lecanii to play a key role in generating the spatial structure of the system was demonstrated using a probabilistic cellular automata model. Finally, a previously unreported one-year lag in the suppression of H. vastatrix by L. lecanii was demonstrated using field surveys. The final chapters demonstrate the potential for agroecological research to inspire more general ecology theory, and for ecology theory to inspire models with implications for the management of agroecosystems. These chapters detail an evolutionary model that demonstrates how a low-connectivity host spatial structure could evolve as an anti-pathogen phenotype; the implications of the self-organization of habitat patches on population persistence; and how recent research on the detection of impending regime shifts might enable the detection of an imminent loss of L. lecanii from the coffee system.