The Community Ecology and Management of Natural Pest Control: Investigating Interaction Complexity and Local Farm Perturbations in Coffee Agroecosystems

Abstract

Agricultural intensification is the principal driver of many of our global environmental crises, including the alarming declines of terrestrial insect populations. To curb its negative impacts, it is imperative that we reduce pesticide use and shift agriculture to sustainable, agroecological pest control methods. Natural pest control (NPC) can reduce pest damage and make agriculture more biodiversity (and people) friendly by conserving natural enemy habitat in and around farms. However, a general approach to natural pest control has proved challenging to develop. More understanding is needed of the community ecology of these complex systems and of the impacts of farm management, both at the local and landscape scale. My dissertation explores key components of these knowledge gaps, investigating the role of interaction complexity, through non-consumptive effects, and the impact of local farm management perturbations on natural pest control communities. I use the coffee agroecosystem and ant-mediated natural pest control as a model system to explore these questions, prioritizing an experimental approach to promote mechanistic understanding. In my first Chapter, I expand on the broader issues above and situate my work in the context of knowledge gaps in this field. In Chapter 2, I conduct an empirical experiment investigating the non-consumptive effect of an aggressive, coffee-foraging ant that removes pests from plants and throws them to the ground. I show how this effect mediates multi-predator interactions and spatial complementarity, and also enhances overall pest damage suppression. I also demonstrate how this interaction can be driven by variation in alternative resources for the non-consumptive predator. In Chapter 3, I model the dynamics of this interaction more generally and demonstrate how non-consumptive effects can stabilize the complex dynamics of spatially coupled oscillators, mediate multi-predator interactions, and enhance pest regulation. In Chapter 4, I explore the broad impacts of disturbance through coffee cutting and renovation. This disturbance disrupted ground cover and coffee plant density, which reduced diversity and shifted the composition of the whole ant community. Arboreal ants were especially affected, with a significant reduction in species richness and abundance for several years, until an eventual recovery five years after disturbance. In Chapter 5, I show how contrasting farm management (high-input sun monoculture vs low-input shaded polyculture) impacts ant habitat through the epiphyte communities on coffee plants, where higher diversity and abundance of epiphytes in the less intensive farm may be driven by differences in the frequency of coffee plant pruning and distance from nearby forest patches. In Chapter 6, I demonstrate the impacts of experimental removal of coffee epiphytes on ant communities and pest control. I show that this intensive practice significantly reduces ant diversity, activity, and abundance on coffee, and ant-mediated pest damage reduction. I also show how this disturbance may interact with climate stress to exacerbate impacts on ants. In Chapter 7, I attempt to synthesize my results and place them in the context of our broader understanding. I explore how these findings illuminate the importance of interaction complexity (non-consumptive effects), which can result in non-intuitive impacts, compared to direct consumptive pest regulation. I also provide greater mechanistic understanding of the impacts of local farm intensification and perturbations, by experimentally testing several types of disturbance on ant natural pest control. I conclude with a discussion of the broader implications and suggestions for future research on these subjects.